Découvrez Les IA Experts
 |
Nando de Freitas | Chercheur chez Deepind |
 |
Nige Willson | Conférencier |
 |
Ria Pratyusha Kalluri | Chercheur, MIT |
 |
Ifeoma Ozoma | Directrice, Earthseed |
 |
Will Knight | Journaliste, Wired |
Profil AI Expert
Non Disponible
Les derniers messages de l'Expert:
2023-03-27 18:32:17 @MarkHaselgrove @lei_zhang_lz Nice :)
2023-03-26 11:46:56 Trigger a hundred papers recreating mid-twentieth century psychology experiments. Do we know if it exhibits blocking and overshadowing yet? https://t.co/KwlAfuD89S
2023-03-23 23:37:39 @mbeisen Plasticine session. Here’s the cerebellum and basal ganglia https://t.co/uKFEhBuIPg
2023-03-23 23:21:15 @mbeisen
2023-03-23 22:30:37 Look what my lab made :) https://t.co/vRzqfejZwJ
2023-03-20 19:08:44 RT @TomNotGeorge: Nice article from @SWC_Neuro highlight my latest paper but, more importantly, a story of how multiple groups can work in…
2023-03-17 21:58:05 RT @mbeisen: The @eLife Board of Directors made the strategic decision to pursue this path &
2023-03-16 12:36:58 @NilsKolling @micahgallen I always thought the rationale for box of frogs was that once you open the lid things start jumping out all over the place :)
2023-03-16 08:48:19 @NilsKolling @micahgallen Where I’m from it’s a box of frogs. Ferrets definitely better!
2023-03-16 08:45:22 @nathanieldaw @micahgallen :)
2023-03-16 08:28:20 @micahgallen This example (unlike some of the other kind-boggling ones) does seem like thoughtless meta-pasting. I think we still have jobs. Phew!
2023-03-16 08:24:40 @micahgallen Also read-out is wrong in two-step. Should be opposite next-trial behaviour on rare vs common, not speed of switching. It’s measure doesn’t distinguish learners.
2023-03-16 08:20:46 @micahgallen Latent learning task also missing the control condition. Might be interesting to ask it what test it would do on the data to discern whether a mouse was using model-based strategy in LL task.
2023-03-16 08:17:49 @micahgallen Three correct. Two dramatically incorrect. As in paradigms do not distinguish the learners but it suggests that they do, and then incorrectly explains how, no?
2023-03-11 20:59:49 RT @GabyMohamady: Frontal neurons fire for particular task states, but seem to remap between tasks. Turns out this remapping reveals a surp…
2023-03-11 06:29:57 This should be totally awesome!! https://t.co/hNhO5twJFv
2023-03-11 06:29:37 RT @LorenzoPosani: The season finale of the mixed vs. modular selectivity debate will take place this year at #cosyne2023 workshops! Jo…
2023-03-10 19:51:48 RT @mbeisen: Please weigh in. I am genuinely interested, because, while there is no perfect solution here, we want to get as close as possi…
2023-03-10 19:06:48 RT @arusbridger: Dear @BBC. In the interests of fairness could you please now ask @Lord_Sugar to step aside from @bbcapprentice . Otherwise…
2023-03-10 08:09:43 RT @prokraustinator: Reference requests that ask you to rate applicant as in the top {1%, 5%, 10%, etc} are hilariously innumerate. Reli…
2023-03-08 06:25:16 This is super cool!! https://t.co/5UN65QG2UT
2023-03-06 19:09:30 @MillerLabMIT The wave seems just like the spike travelling down the axon to me..
2023-03-06 19:08:34 @MillerLabMIT And easy to infer the wave from individual recordings if you also measure xy (so long as it is stim aligned).
2023-03-06 19:07:32 @MillerLabMIT Wait what - that’s cheating no? You only see the wave because you have added new dimensions to your measurements (xy location) not because you have recorded lots of things together. If you didn’t measure xy you would just see lots of people going up and down.
2023-03-05 18:50:42 @MillerLabMIT Got it. Thanks earl :). I guess a spike is an emergent property of ions and then electrons shuffling around too so one person’s emergence are another person’s cogs :)
2023-03-05 10:00:00 CAFIAC FIX
2023-03-02 22:00:00 CAFIAC FIX
2023-02-27 01:00:00 CAFIAC FIX
2023-02-20 11:51:52 @AnnDuan2 @CortexClub @flickerfusion @ArminLak @paqio @summerfieldlab awesome - but I think I'll be at SWC Weds (my day to take kids to school so will likely be in London :( )
2023-02-20 11:51:07 RT @AnnDuan2: Giving a talk at Oxford on Wednesday (thanks for the invitation @CortexClub), chatting about the new questions we are pursuin…
2023-02-15 18:17:22 @ScottishWaddell @PhilCorlett1 They changed the name to platform grant, but it was effectively an refunding bid because only existing centres were allowed to apply. We didn't get it. Other centres did.
2023-02-15 14:07:43 @jtdudman @PhilCorlett1 Thanks Josh!!! Hopefully we’ll manage to find some support elsewhere and keep trundling on with the same kind of nonsense:)
2023-02-15 07:53:15 @neurovidaurre @PhilCorlett1 None given yet. We’ll talk to them in about a month
2023-02-14 19:58:08 @gershbrain @PhilCorlett1 Thanks Sam :)
2023-02-14 19:51:56 @PhilCorlett1 Yep we were pretty surprised. Tough to imagine what else we could have done. We thought the combination of U.K. biobank, FSL, cross-species stuff was pretty unique. Apparently we didn’t make a good enough case.
2023-02-14 19:47:54 @PhilCorlett1 Our centre in Oxford also just got defunded by Wellcome. Not as bad news as CBU as core grant was much smaller (not covering researcher salaries) but nevertheless will make life a lot harder.
2023-02-14 19:42:17 RT @PracheeAC: Roses are red Violets are blue Give authors the power As @eLife will do ICYMI, the new model is live! https://t.co/VZdAW…
2023-02-14 14:29:02 @SWC_Neuro @GatsbyUCL @BrainBee_UK @LondonMetUni @_JasvinKaur @BackyardBrains @BeatrizSimG
2023-02-13 20:28:28 RT @ShirMarkLey: Israel is fighting for its democracy. I cannot believe that I should shout for women rights and democracy in Tel Aviv and…
2023-02-11 20:51:14 @leslievosshall Yippee!! So pleased you like it :)
2023-02-11 20:50:42 RT @leslievosshall: @behrenstimb I really enjoyed this experience! It's an excellent paper and it was interesting to see the reviewers' tak…
2023-02-11 18:14:32 @singletonion @leslievosshall Yes you can mint a version of record at any time after the first review. Then it goes into Pubmed etc.
2023-02-11 18:13:29 @NeuroecologyLab @EFouragnan As soon as you declare you are finished with the paper, we mint a version of record and it goes in Pubmed.
2023-02-10 16:20:59 @cshperspectives We would like to continue to support them, but not immediately obvious how.
2023-02-10 16:20:47 @cshperspectives The particularly tricky thing with Review Commons is not coordinating the different RPs. It is that they still would like a journal to accept (rubber stamp) the paper. Previously we have been one of those journals. But the new model does not rubber stamp things.
2023-02-10 10:45:44 @AMartinezArias @pavi_narayanan @ReviewCommons @eLife No! It always was that. Now it is not. The only confusion is wrt to review commons that cannot use the new model because they need a selective journal to rubber stamp. This conversation is totally mad :)
2023-02-10 10:44:11 @AMartinezArias @ReviewCommons @eLife I totally agree with you that there is a pickle about how new eLife interacts with review commons. Everything else seems really really clean to me. I think the new model is extremely pure. Just put it on biorxiv and we will review and assess it for you.
2023-02-10 10:41:15 @AMartinezArias @pavi_narayanan @ReviewCommons @eLife Nothing wrong with review commons at all. But they need a journal to publish their papers after review. They need the rubber stamping. That's their model. Our new model does not offer that.
2023-02-10 10:40:02 @AMartinezArias @pavi_narayanan @ReviewCommons @eLife In some subject areas they definitely are. Review commons is strong in cellular biology, but has no visibility in some other areas. eLife definitely has a wider selection of reviewers in neuroscience, for example.
2023-02-10 10:38:04 @AMartinezArias @ReviewCommons @eLife If the paper gets past the selection and is accepted, we will publish all of the reviews.
2023-02-10 10:37:23 @AMartinezArias @ReviewCommons @eLife No - I don't think you are understanding what I am saying. I will try one last time. You cannot bring review commons papers to the new model. You can only bring it to the old model. The old model is *selective* after peer review, just as it always was.
2023-02-10 10:24:47 @pavi_narayanan @AMartinezArias @ReviewCommons @eLife @mbeisen No because they can still go through the traditional model. See tweets further down thread.
2023-02-10 10:23:18 @pavi_narayanan @AMartinezArias @ReviewCommons @eLife In the meantime, we are still offering review commons papers to go through the old model, where we can offer them what they need (a journal rubber stamping). So we can be an affiliate journal that way.
2023-02-10 10:22:23 @pavi_narayanan @AMartinezArias @ReviewCommons @eLife If reviews and assessment are the final product then it is important we produce high quality reviews and assessments. We cannot guarantee that with outside suppliers, so we need to work out how review commons etc. plays with the new model. We don't know yet.
2023-02-10 10:19:41 @pavi_narayanan @AMartinezArias @ReviewCommons @eLife Not like review commons because review commons thinks you get reviews *then send it to a journal for rubber stamping and curation (which journal it ends up in really matters)*. We think that reviews and assessment *are the final product*.
2023-02-10 10:17:20 @AMartinezArias @ReviewCommons @eLife I am totally unsure what the confusion is. Review commons position is that preprints plus reviews are not enough. They need the rubber stamp of a journal "accept" decision. We cannot offer that under new model. So we offer it under old model.
2023-02-10 10:16:01 @AMartinezArias @ReviewCommons @eLife eLIfe old model was a traditional journal. eLife new model is not. eLife new model publishes reviews for every paper we review. Review commons needs a journal to publish its papers. We are still offering to be a journal via the old model for review commons.
2023-02-10 08:57:25 @AMartinezArias @ReviewCommons @eLife And yes we will charge for publishing just like we always did.
2023-02-10 08:56:47 @AMartinezArias @ReviewCommons @eLife Together they will decide whether to publish or reject ask for revisions etc. We are **not** handling any review commons papers under the new model right now. Just via the old-fashioned route.
2023-02-10 08:55:29 @AMartinezArias @ReviewCommons @eLife Sorry - I thought I had answered this question. It is just the same as before. It will go to editors who will decide whether it meets the arbitrary eLife "bar". The editors might reconsult the reviewers, or might sometimes even consult new reviewers. 1/2
2023-02-10 08:18:26 @CellTypist Or if they want to, they can take these reviews and use them in a submission to Nature. Their call :)
2023-02-10 08:11:54 @AMartinezArias @eLife @ReviewCommons Review commons papers are currently handled in the traditional way (not in the new model). We’re in conversations about what to do with them going forward.
2023-02-10 08:09:48 @CellTypist It already is. But if authors want it in Pubmed etc. They just have to say so. That mints a version that will not be further updated.
2023-02-10 06:34:15 @FailledN @eLife I guess your comments are made as snarks. I’m happy for you to snark at me. There are lots of things that need to happen to break this cycle and I’m proud that I am investing a lot of energy in one of them.
2023-02-10 06:14:48 @FailledN @eLife Paper would have got into Nature. And we can do it much faster. We also offer an author-controlled experience which must be better. So I’m hopeful. Yes. Definitely my lab are more interested in elife now.
2023-02-08 13:53:50 @PlantEvolution @GilCostaDesign https://t.co/yrEf4P6Arv
2023-02-08 11:47:32 RT @summerfieldlab: please submit to CCN!! (dates 24-27 August in Oxford). Submission deadline 31st march. Also, register early - reduced p…
2023-01-30 01:00:00 CAFIAC FIX
2023-01-14 18:47:08 RT @adredish: Cool new paper by lab alumnus @zebkDotCom: https://t.co/KFFqvUmwSk. Some corroborating data: same data is in https://t.co/rIw…
2023-01-14 13:39:47 RT @KiaNobre: Mark Stokes (@StokesNeuro), RIP You enriched us with your fortitude and gentleness. You changed our scientific views with you…
2023-01-13 18:32:14 @zebkDotCom
2023-01-13 17:00:30 Some cool new ideas about composition and replay from Zeb Kurth Nelson et al.. https://t.co/XmMOxWwkAk
2023-01-13 05:39:16 RT @TheBrunoCortex: VISUAL NEURO JOB OPENING: The University of Oxford's Dept.Phys.Anat. &
2023-01-12 06:49:08 RT @Samfr: Which means tuition fees will have fallen roughly a third in real terms since they were put up by the coalition govt.
2023-01-10 17:14:23 @LHuntNeuro I haven't been in that building for approx 20 years. Might get PTSD.
2023-01-10 17:13:47 @LHuntNeuro :)
2023-01-10 17:09:44 This is excellent news. Unfortunately it means that computers will not be allowed at the conference, but on the plus side, if you come wearing your sword the organisers will be compelled to supply you with free beer. @LHuntNeuro, will you be supplying the coloured carnations? https://t.co/8LSaBlJR4y
2023-01-05 14:08:28 @DrDanielCM Ha - I remember that talk too. It seemed logical at the time, but I can't remember why!
2023-01-04 23:05:31 @Franklandlab Yes yes I’m sure it’s useful, but it doesn’t stop me feeling grumpy:)
2023-01-04 18:54:11 I hate writing new talks. It takes me about a week. I spend the first 3 days writing completely the wrong talk, being so dissatisfied with it that I procrastinate on twitter or elsewhere. Apparently only after this failed process can I begin to write something coherent.
2023-01-04 14:04:21 @TrackingActions Ha - awesome. This is for an example in a talk (human only). But such a fun idea :)
2023-01-04 11:11:28 I have had a sneak peak at the proposed experiments on this project and they are https://t.co/6PeqQmtbD0
2023-01-04 11:09:22 RT @erlichlab: The deadline to apply is Jan 8!!
2023-01-04 07:02:39 This is really fun :) https://t.co/A5AN3mTy7Q
2023-01-03 19:08:37 @crewilson @CousinAmygdala What do you expect to happen? All the good stuff is 20 years old. Can't keep riding Matthew's coat-tails forever.
2023-01-03 18:00:33 RT @mirandarwolpert: Senior Data Specialist @wellcometrust Salary: £68k Closing date: 22nd January Lead our strategy to fund cutting- e…
2023-01-03 18:00:15 @jryandx Awesome - that looks perfect. Thank you!
2023-01-03 17:23:34 What is state of the art for reconstructing images or video from fmri data using ANNs?
2023-01-03 16:51:02 @LHuntNeuro @CousinAmygdala :) Just hoping the junk mail filter steps up its game in 2023, and takes them off my hands. Happy new year mate.
2023-01-03 12:49:34 @CousinAmygdala These are all from my mum. Yesterday was a busy day for her.
2023-01-02 09:06:13 RT @paulisci: A List of Predictions Made in 1923 About 2023
2023-01-01 15:47:20 Any topology experts follow me? https://t.co/EwQdPx0uI9
2022-12-31 14:12:29 @AnnaSchapiro This is so good to hear. All fingers and toes crossed for a full recovery!!
2022-12-30 19:09:57 @KordingLab But what about the fun. I come to Twitter for science chat, but also for comedians journalists, sport, non-sci friends. So depressing to lose that in favour of a place dominated by science debate, interesting though that is. Mastodon is far too worthy atm.
2022-12-29 18:04:22 @derdikman You have dealt with this so beautifully. It was a clumsy first tweet but your responses and follow-ups have done you proud. If people knew you they would have seen the original tweet as a generous tweet about the paper, as it was intended.
2022-12-28 08:04:57 RT @a_m_mastroianni: Recently I wrote a post suggesting that peer review doesn't work, and then some weird things happened. A tenured profe…
2022-12-24 17:29:37 RT @TheBarmyArmy: https://t.co/0CDbqagWep
2022-12-23 18:54:20 RT @dan_grey: Extraordinary set of charts in the FT today. Twelve years of Tory misrule has left us sicker and poorer than our peers. Utter…
2022-12-21 19:20:32 RT @GavinBarwell: To give an idea of the scale of the impact, the March 2022 Budget raised taxes by £46 billion. In broad order of magnitud…
2022-12-18 09:19:54 @StokesNeuro Such a beautiful message. You’ve been a beacon of thought, creativity and strength. Sending love to you and your family.
2022-12-16 19:11:55 RT @dorrell_will: New pre-print!! With Maria Yuffa &
2022-12-16 09:47:34 RT @dileeplearning: Location, encoded by place cells, is important for animals, humans, … and airplanes. Check out our new work: "Space i…
2022-12-13 05:53:54 RT @AdamCSharp: My favourite ever gritter names… 10. David Plowie 9. Nicole Saltslinger 8. Spreddie Mercury 7. I Want to Break Freeze 6. G…
2022-12-08 13:00:00 CAFIAC FIX
2022-12-07 08:00:00 CAFIAC FIX
2022-11-14 08:16:31 RT @TrackingActions: @SuryaGanguli @behrenstimb @KordingLab @PessoaBrain @SandeepKishor13 @NicoleCRust @snsf_ch Well… EPFL is a wonderful p…
2022-11-14 06:21:05 @SuryaGanguli @KordingLab @PessoaBrain @SandeepKishor13 @NicoleCRust Weird. Seems different in Europe. Wellcome,ERC,Gatsby, IST, Max Planck all have big investments in theory.
2022-11-13 20:48:45 @hjhope Have done :)
2022-11-13 20:35:49 @hjhope Is there a chance we could chat briefly about what would be possible/useful?
2022-11-12 21:37:56 @hjhope I realise Wellcome accept preprints even without peer reviews in grant apps :)
2022-11-12 21:37:06 @hjhope Hi Hannah. Is it possible for Wellcome to issue a statement saying this. Lots of people are worrying unnecessarily :). Would also be good to clarify how, in grant apps, Wellcome will treat reviewed preprints before authors have opted to cast a version of record.
2022-11-12 19:23:10 RT @nikasamborska: Ukrainians will need to create a new word to capture the emotions we are all feeling right now
2022-11-12 18:08:26 Amazing achievement. Technical. Social. Scientific. https://t.co/ac6E8jZ8eB
2022-11-12 18:07:23 RT @IntlBrainLab: BIG DATA RELEASE We are beyond excited to announce the release of our Brain Wide Map of neural activity during decisi…
2022-11-11 22:42:06 Ok I made onehttps://t.co/XT5ceG2NCp
2022-11-10 19:18:25 RT @drkjjeffery: We’re looking for in vivo rodent electrophysiologists interested in complex behaviour to come and work in beautiful Glasgo…
2022-11-10 05:33:46 RT @hjhope: Sort of thought this went without saying but....As a member of the board, @wellcometrust supports the transition eLife is mak…
2022-11-09 21:23:45 RT @brody_lab: For the new model to work for all of us, we all have to help it. I’m going to be encouraging (but not requiring) everyone i…
2022-11-09 12:18:30 RT @MarWimber: Here's the paper &
2022-11-09 12:18:19 RT @MarWimber: @eLife @CKerren @sandervanbree @b_j_griffiths My experience with @eLife has been positive throughout, as reviewer, guest edi…
2022-11-09 12:18:10 RT @MarWimber: Paper accepted in @eLife! Stellar work by @CKerren with dream team @sandervanbree &
2022-11-09 06:50:28 RT @SnoozeInBrief: He lasted 0.37 Kwartengs.
2022-11-08 22:02:22 Good luck America.
2022-11-08 15:40:55 @EngertLab Huh? This is all just either Pavlovian approach or Pavlovian repurposing of a "Swim-through" reflex isn't it? The only one that might be different is if it had to swim 3 times through the hoop before surfacing for reward. Was that the rule?
2022-11-07 17:47:14 @gershbrain where/what's the other one?
2022-11-07 16:27:05 @Dick_retired @Timothy_Bussey https://t.co/8aIZsWMuMI
2022-11-07 16:25:46 @Dick_retired @Timothy_Bussey I tried to get on fediscience this morning. So far no joy. They need to confirm I am a scientist or something :)
2022-11-04 18:33:05 RT @Anne_On_Tw: Many years of work, published! Congratulations @eckstein_maria @SarahMaster17 @WilbrechtL @ccnlab and collaborators not on…
2022-11-03 10:18:43 @AMartinezArias @KamounLab @ThorstenLangner @drdevangm @PlantEvolution @eLife eLIfe is like an oil tanker. It takes a long time to change direction and you can't change too much at a time :)
2022-11-03 10:18:00 @AMartinezArias @KamounLab @ThorstenLangner @drdevangm @PlantEvolution @eLife Sorry. Didn't mean to come across as defensive. Just trying to separate one issue that we have thought carefully about, from another which we have not yet. Much of the criticism of our policy has been about triage, which is the only bit we haven't dealt with yet :)
2022-11-03 09:23:44 @KamounLab @ThorstenLangner @drdevangm @PlantEvolution @eLife I guess I am saying that you can criticise our triage process, and you should. We agree. But at the same time, you can celebrate our attempt to change what it means to publish a paper.
2022-11-03 09:23:02 @KamounLab @ThorstenLangner @drdevangm @PlantEvolution @eLife We agree with you that triage needs a revamp, and transparency will def be a factor in that, but I think there has been some confusion about the relationship with the current move, which is really not about transparency at all.
2022-11-03 07:04:32 West London is a river https://t.co/7J0rSGMqOq
2022-11-03 06:27:08 RT @fmrib_steve: $65k.In Oxford it is $39k, with a cost-of-living salary increase this year of less than $2k. This isn't sustainable.
2022-11-02 06:33:32 RT @TheNeuro_MNI: Winner of a 2022 Neuro - Irv and Helga Cooper Foundation Open Science Prize, @neuromatch fosters inclusive global interac…
2022-11-01 21:13:26 RT @nachopolti: New paper @eLife with co-first author @NauMatt!Hippocampal activity reflects timing behavior &
2022-11-01 21:11:17 .@ilyasut could learn a thing or two from @dendritesgr https://t.co/dF7ouIySR0
2022-11-01 06:41:27 RT @SuryaGanguli: Our new paper @NeuroCellPress "A unified theory for the computational and mechanistic origins of grid cells" lead by Ben…
2022-10-31 08:14:54 RT @nick_kapur: Mundane Halloween is back!Japan does Halloween better than anyone, but best of all is the annual "mundane Halloween" cont…
2022-10-31 08:14:43 RT @nick_kapur: "Person whose skeleton is being estimated by machine learning"https://t.co/PnjxqVHoOv
2022-10-28 12:57:44 @drdevangm @PlantEvolution I'm glad @PlantEvolution returned home full of hope. @mbeisen and I returned home full of Gin. Was v nice to see you guys!
2022-10-28 08:33:29 @mbeisen I am still drunk.
2022-10-28 06:41:04 RT @TPVogels: We find ourselves in the curious position to have very few non-African applicants... Usually it's about ~100 applications fro…
2022-10-27 08:13:15 RT @MIcheleABasso1: Right on Marge!!Professor Livingstone’s personal statement about recent concerns over animal research | Harvard Medica…
2022-10-27 07:59:33 @NeilBurgess10 @drkjjeffery @neuralengine That doc I sent was the first one I think. Can dig out second one at some point. Have to go to elife meeting now though to see if we can resolve all these issues
2022-10-27 07:58:39 @NeilBurgess10 @drkjjeffery @neuralengine Trance = accept. Weird
2022-10-27 07:58:01 @NeilBurgess10 @drkjjeffery @neuralengine We are actually confusing two different trials. Sorry. First one triage plus guaranteed acceptance. Authors could withdraw. Second one no triage. There was a trance decision. But all reviews posted on biorxiv. No gender differences in people opting for either trial.
2022-10-27 07:29:59 @NeilBurgess10 @drkjjeffery @neuralengine Papers were not rejected after peer review. Authors knew we would post all reviews.
2022-10-27 07:28:17 @NeilBurgess10 @drkjjeffery @neuralengine No.
2022-10-27 07:21:26 @drkjjeffery @neuralengine @NeilBurgess10 https://t.co/EONCBOX9GT
2022-10-27 07:19:51 @drkjjeffery @neuralengine @NeilBurgess10 No this was a trial where all reviews were made public in exchange for guaranteed review (no triage). Everything we reviewed in the trial was accepted. Negative reviews published.
2022-10-27 05:33:59 @drkjjeffery @neuralengine @NeilBurgess10 We will be monitoring this but evidence to date suggests it isn’t true. Given a choice in our trial, women were no less likely to opt for public peer review than men. The only demographic less likely to prefer public peer review were late (compared to early) career researchers
2022-10-26 17:40:07 @AllenNeuroLab @stevenmweisberg @smfleming @eLife I **promise you** we’ll get to triage!!! There are more radical solutions than this on the table (as well as less radical ones)
2022-10-26 17:03:51 @AllenNeuroLab @stevenmweisberg @smfleming @eLife There is a real philosophical difference here. The new system is now like New York Times Film Reviews. You can't tell if the film is good just because the review is there. You need to read the assessment. But like NYT, we will still have to choice who we review, causing problems.
2022-10-26 17:01:08 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Some people are definitely shouting. Maybe not you :) We aren't moving peer review decisions to triage decisions. We are getting rid of the notion of acceptance being the goal.
2022-10-26 16:17:11 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Triage is on the agenda. We understand it's an issue. We will also try to figure out a way to do that better. In the meantime, I would be happy if people focused their criticisms on what we are doing with peer review.
2022-10-26 16:16:22 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We have been doing triage behind a barrier for a decade as do almost all journals. Now we change something totally different and everyone shouts at us about triage. It is just madness. It's like I told you I am moving house and you shouting at me because you don't like my work
2022-10-26 16:14:47 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We practically have to handle 7-900 papers a month in a relatively streamlined way. So any changes we do need to be gradual.
2022-10-26 16:13:00 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Not sure I really parse this message, but we can't change everything at once. eLife is like a massive oil tanker that takes a long time to turn. There are 650 editors who all have different views and practices and habits.
2022-10-26 15:23:52 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We know triage is a random and biased and hated thing. Can people let us try to solve one problem at a time please.
2022-10-23 19:58:08 @BorisLenhard @eLife https://t.co/a3SJJfUHqs
2022-10-23 19:39:39 @LilaDavachi @EAHellerPhD @PracheeAC @eLife Just upping this here as well as in DMs because I think people have missed it. Just noting that the more is a concrete evaluation as well as the peer reviews. We should def chat more about this though. https://t.co/a3SJJfDEos
2022-10-23 18:43:12 RT @JACoates: Can we just talk about the reality of peer review for a second because so many scientists here seem completely oblivous to th…
2022-10-23 18:39:38 @LilaDavachi @PracheeAC @eLife Hey Lila - I don't get what you are saying here, but we are really keen to understand the potential sources of bias in this new system that are different from the old. Is there any chance you could unpack it to explain more? Could be in DMs or email if easier
2022-10-23 18:37:33 RT @patrickmineault: Brains are cool! They are literally what make us, us! We should study them more, for their own sake! 6/6
2022-10-23 12:57:44 RT @LMT_Spoon: At the risk of wading into this debate very early in my career and appearing naive about the realities of scientific publish…
2022-10-23 12:57:34 RT @albertcardona: Wonderful to see an early career scientist with such a clear understanding of scientific publishing and what @eLife is t…
2022-10-23 10:19:39 @niallpboyce But how is that different to the current system? You guys are not transparent. You make mistakes. You publish stuff that doesn't replicate. You are entirely beholden to reviewer opinion. We are just saying this reviewer opinion is not binary. That's the only difference.
2022-10-23 09:50:25 @generegulation @eLife You haven't understood then. https://t.co/a3SJJfDEos
2022-10-23 09:45:41 @aazadmmn Haha - No I am just being childish. No subtext at all. Literally referring to the colour of a dead and living brain and the appearance of figures in ML papers. No depth at all. Just silliness. Trying to avoid any meaningful conversation about AI and neuro.
2022-10-23 09:41:16 @aazadmmn They don't seem to have a regular colour in my experience. Definitely not boxes though. When I see them they are usually more like rectangles with arrows pointing out of the ends. Sometimes trapeziums. Occasionally circles. Always arrows though.
2022-10-23 09:36:37 @aazadmmn Even the dead brain is at most a browny-grey box. Much easier. When it is up-and-running it is pinky-red. Piece of cake.
2022-10-23 09:34:04 @generegulation @eLife This is a really strange take. In either case the journal is endorsing the reviewers. Previously this was only indirect. Now direct. That is exactly what the "journal" brings. Endorsement of the assessment (not the paper).
2022-10-23 09:32:24 @generegulation @eLife The journal can put its reputation at stake by endorsing the reviews and assessment!!!
2022-10-22 18:02:14 Figuring out how the brain works seems hard enough to me. I'm happy to let someone else solve AI.
2022-10-22 15:23:01 @mazefire56 :) @jcrwhittington @dorrell_will
2022-10-22 15:12:48 @TonyZador I guess the first autoencoder that solved the variational inference problem was the Helmholtz machine (Hinton, Dayan, Senjowski) - a model of the visual system, trained by the wake-sleep algorithm. Although perhaps Zhaoping Li had an olfactory version earlier.
2022-10-22 14:53:19 @TonyZador Transformers are an extropolation of attention models ("Attention is all you need"), which were definitely inspired by human vision. Prob not neuro, but def psychology. eg https://t.co/WMpSNeoKNx
2022-10-22 11:12:00 @lam88324 @RafaCarazoSalas @micahgallen @vsbuffalo Ha - It's been 48 hours - give people a chance!! Almost every institution in the scientific world has signed a statement that they won't judge by IFs but they still do because there is no alternative. Let us try to build one, and join on this journey :)
2022-10-22 08:50:42 @PlantEvolution @fishspeciation @eLife @mbeisen I don't understand why people are assuming the worst from a charity run by people trying to improve things. In general with eLife ->
2022-10-22 08:47:18 @fishspeciation @PlantEvolution @eLife @mbeisen As far as I know we have never yet declined a fee waiver request. There is your data right there :)
2022-10-22 07:53:56 RT @DanGarrahan: The Brexit effect film is free to view. No remoaning - just cold, hard facts clearly explained by @FT writers and interv…
2022-10-22 07:09:18 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We totally agree with you that triage is a problem, and transparency is important. It’s on the agenda. One step at a time.
2022-10-22 07:08:25 @drkjjeffery @joshdubnau @WesterfieldJM @eLife Instead we are trying to change what it means to publish. To change the output of science evaluation to be the evaluation itself. We think this brings many benefits as laid out in the editorial. But it isn’t about transparency.
2022-10-22 07:06:28 @drkjjeffery @joshdubnau @WesterfieldJM @eLife This misunderstands what we are doing. One reason for public evaluation is transparency of decision. We agree this is important (and triage is a big part of the decision) but this is not what we are doing here.
2022-10-22 06:14:36 @BalsdonTarryn Correct stats are as follows ->
2022-10-22 06:12:24 @BalsdonTarryn Hi. This is inaccurate. Approx 50% of papers reviewed papers are rejected. Your stats are for papers that are invited for revision. We have always tried to ensure that after we invite a revision chances are great. Please correct. Thanks.
2022-10-22 06:09:00 RT @GatsbyUCL: Upcoming Faculty Openings!We intend to recruit one principal investigator in #TheoreticalNeuroscience and one in #Machine…
2022-10-22 06:04:58 @adredish How’s this for your CV.“This paper reports the landmark discovery that mouse hippocampal neurons rapidly play out different possible futures during a decision. This suggests a new role for hippocampus and a new mechanism for choice. The supporting evidence is compelling.”
2022-10-21 21:44:40 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell Oh I see. I think for the next 3 months triage will likely be similar as editors get used to the new system. A problem with a massive machine is that it takes a long time to turn.
2022-10-21 21:42:01 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell We will def publish more papers. We hope to *review* more papers too.
2022-10-21 21:41:21 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell What do you mean A vs B route?
2022-10-21 21:40:38 @ElDuvelle @AheadOfTheNerve It will be indexed in Pubmed so unfortunately it will get an impact factor. I wouldn’t be surprised if Thompson do not even recognise that there has been a change.
2022-10-21 21:38:46 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We also have a very high-touch system where many editors see each decision. I think this is important when reviews are public. Lots of layers of scrutiny. Hopefully this means we are less likely to screw it up.
2022-10-21 21:36:30 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We are a large scale trusted journal, and we are pivoting to public evaluation. That makes us unique I think.
2022-10-21 21:35:43 @drkjjeffery @joshdubnau @WesterfieldJM @eLife Nothing that others couldn’t do in time. But we have a massive head start - a huge machine of 650 editors who are field leaders committed to this. And we will do it all publicly with a serious attempt at consistency across the journal, and an eye to reviews that are useful.
2022-10-21 17:08:27 @AllenNeuroLab @stevenmweisberg @smfleming Yes it is not a peer reviewed journal. The papers are *already* published on biorxiv. We are not selecting which to publish. We are attaching a public assessment to some of them.
2022-10-21 17:05:42 @AllenNeuroLab @smfleming In the same way new york times decides which films to review. Yes there are biases here which we should worry about, but critically you can't tell whether the film is good without reading the review. It is very different when the decision itself does not signal quality.
2022-10-21 14:58:37 @RafaCarazoSalas @micahgallen @vsbuffalo @eLife The only way eLIfe dies is if it stops innovating. What else is it there for? Plenty of prestige journals.
2022-10-21 14:57:23 @ScottBrincat @vsbuffalo I think the reception has been mostly good. It's difficult to know what aspects to focus on, and different for different sectors. We decided to focus on the overall mission to start with then shift the focus to the details on the following days. We're moving slow :)
2022-10-21 14:11:29 RT @LnccBrown: as a long time senior editor for @eLife I (MJF) thought I'd weigh in. When I first heard about the plan in its initial conce…
2022-10-21 14:09:14 @joshdubnau @WesterfieldJM No. @elife badge will say nothing about the quality of the paper. It will promise a high quality review. That's the whole point. The paper is already pubilshed before we even see it. This whole thread misses this critical point.
2022-10-20 21:12:56 RT @JPox14: eLife is a force for good
2022-10-20 21:09:26 RT @mbeisen: To enhance their clarity and consistency, these assessments draw on a controlled vocabulary that provides editors with a semi-…
2022-10-20 21:09:16 This is what it will look like. With embedded reviews https://t.co/ihrQDnopTT
2022-10-20 21:08:46 RT @mbeisen: Now that I've had a chance to see peoples' questions about @eLife's new publishing model, time for a thread for a thread with…
2022-10-20 20:16:26 RT @neuralreckoning: @tyrell_turing @behrenstimb @eLife This is exactly right, the key question for the big bet eLife is making. Let's not…
2022-10-20 20:15:55 @tyrell_turing @eLife :) hoping I can write that assessment for real one day…
2022-10-20 20:11:59 @tyrell_turing @eLife Already Wellcome/HHMI asks for them, but the applicant has to write them. Imagine how much better if they are written by independent experts!!!
2022-10-20 20:10:55 @prokraustinator @tyrell_turing @eLife Because nobody writes clear concise summaries that hiring committees could understand. We (the applicant) have to write these for Wellcome grants etc. imagine how much more impressive if the reviewers write them for you.
2022-10-20 20:08:25 @DanHandwerker @tdverstynen @gershbrain @eLife Yes we definitely can detect some fraud. We could retract in this situation but I suspect we won’t as the paper is on biorxiv and we will want to highlight that it is fraudulent. So we will likely cover the paper in red ink.
2022-10-20 20:01:54 @adredish Also kinda weird to quote-tweet this like you are showing me up??? I have been an editor for 15 years. I know the difference between fraud and error.
2022-10-20 19:59:46 @adredish We may not detect fraud. It often can’t be identified by submission alone. but we will do our very best to find flaws, which can, and highlight them.
2022-10-20 19:58:15 @adredish I was making exactly that distinction.
2022-10-20 19:49:29 @tyrell_turing @eLife Sound good enough on your CV?
2022-10-20 19:48:52 @tyrell_turing @eLife Richards et al. is a landmark paper that uncovers the mechanism underpinning gradient descent in the brain. The evidence supporting the proposed mechanism is exceptional.
2022-10-20 19:45:26 @tyrell_turing @eLife The thing that will go on your CV is no longer the citation. It is the assessment. https://t.co/a3SJJfDEos
2022-10-20 19:12:55 @adredish @tdverstynen @gershbrain @eLife ?? I think you’re saying exactly the same as me. All we are saying is that when we know papers are flawed we should say so.
2022-10-20 19:06:17 @adredish We will provide a clear concise assessment that will judge qualityhttps://t.co/a3SJJfUHqs
2022-10-20 18:56:21 @tdverstynen @adredish @gershbrain @eLife It will not be possible to misunderstand. It will be very prominently placed under the abstract. This argument has been used against biorxiv but it turns out people are actually pretty clever.
2022-10-20 18:52:00 @adredish @tdverstynen @gershbrain @eLife All of that independent of vor (which is for indexing reasons)
2022-10-20 18:51:03 @adredish @tdverstynen @gershbrain @eLife Yes. We will post reviews on biorxiv and also host the preprint (displayed a lot more nicely) on our website with reviews embedded
2022-10-20 18:46:57 @adredish @tdverstynen @gershbrain @eLife Scientific results are not binary believe/no-believe entities. They come with different strengths of evidence. We lay this out clearly so everyone can understand.
2022-10-20 18:44:45 @adredish @tdverstynen @gershbrain @eLife We will write clear concise statements telling them exactly which to believe and what the strength of evidence is. Much better than current system. You can see how it works here:https://t.co/a3SJJfUHqs
2022-10-20 17:41:21 @payampiray @tdverstynen @gershbrain @eLife No he's saying the same as me. He wants to reduce the number we triage (currently 2/3), so that we review more papers.
2022-10-20 17:20:28 @jmourabarbosa @tdverstynen @gershbrain @eLife And check Jorn's great blog about it here:https://t.co/9thr3I3T51
2022-10-18 07:15:39 RT @summerfieldlab: I think it's probably time to share this...I wrote a book. It's coming out in dec 2022 (in theory). It is intended to h…
2022-10-18 05:57:55 Sean is totally awesome. Just starting a new group in Bristol. This will be a really really cool PhD. https://t.co/Mh3hJoWEKI
2022-10-17 12:05:27 @SaadJbabdi @richardosman @thebookferret So sorry mate. Thinking of you and pixel.
2022-10-15 05:41:15 @JSchleiderPhD Here’s a list of some of mine https://t.co/6Q1YgqNAvH
2022-10-14 05:08:17 The big limit that is not overcome-able is the implied repetitive experimental design. Still a huge opportunity though. So good to see such strong data supporting a claim like this for once. https://t.co/VXYyLgvL9E
2022-10-14 05:04:57 RT @fMRI_today: Stunningly well done study. Direct neuronal activity imaging with MRI. They nailed it. Practical limits but all manageable.…
2022-10-14 03:46:44 Looks great fun :) https://t.co/u9Xf9vUgHO
2022-10-13 19:22:20 RT @libankerDC: So hear me out: NO MORE RECOMMENDATION LETTERS FOR ANYTHING BUT JOBS. No letters for: internships, study abroad programs,…
2022-10-12 19:33:06 ffs https://t.co/i2Ztkp9aEV
2022-10-12 19:29:56 RT @Cajal_Training: #Brain #robot competition at @SWC_Neuro! Apply to the Cajal #Online NeuroKit #Experimental #Neuroscience Bootcamp to bu…
2022-10-12 15:31:40 Amazing :) https://t.co/OK9u4hG1CS
2022-10-12 15:31:10 RT @bradpwyble: If your argument for the current scientific literature is that publishers add value by copyediting, no humans looked at thi…
2022-10-12 05:19:00 RT @payampiray: I'll be recruiting PhD students for Fall 2023. Take a look at the lab website and the FAQ for prospective students:https:/…
2022-10-11 17:21:02 RT @doellerlab: Group leader position (W2
2022-10-08 07:23:23 RT @GeorginaMacKen4: Neuroscience funding opportunities are available @wellcometrust through our:Discovery Research Funding SchemesTar…
2022-10-08 06:18:34 RT @tyrell_turing: PSA for scientists:Even if a journal's submission instructions say to place figures at the end of the paper *ignore th…
2022-10-07 12:12:38 This is fab :) https://t.co/T2WybvB4Sn
2022-10-07 12:12:18 RT @GabyMohamady: My neuromatch talk is up on crowdcast!We find mFC neurons arranged into multiple, spatially anchored, task-space rings.…
2022-10-07 05:46:02 @steliebe @matthijs_pals @mackelab @humansingleunit Congrats on a fantastic paper.
2022-10-07 05:44:17 RT @steliebe: I don’t remember what life was like before starting this project years ago, but somehow our brains track the order of events…
2022-10-06 21:35:40 @StanDehaene @GabyMohamady @jcrwhittington @timoflesch @summerfieldlab @SaxeLab In motion trials - sorry
2022-10-06 21:28:44 @StanDehaene @GabyMohamady @jcrwhittington @timoflesch @summerfieldlab @SaxeLab I’d love to show you the factorisation in @GabyMohamady’s data. I don’t know if yours will be similar, but it might give us some ideas for discussion.
2022-10-06 21:27:02 @StanDehaene @GabyMohamady @jcrwhittington @timoflesch @summerfieldlab @SaxeLab But that aside, the reanalyses does show a factored rep (and intriguingly in the current discussion that the factors occupy different neuronal populations)
2022-10-06 21:25:54 @StanDehaene @GabyMohamady @jcrwhittington @timoflesch @summerfieldlab @SaxeLab Another subtlety in that task is that the task variables are not unite factored in the behaviour because choice and motion are always in the same direction in choice trials.
2022-10-06 21:24:42 @StanDehaene @GabyMohamady @jcrwhittington @timoflesch @summerfieldlab @SaxeLab In particular if you look in the supplement you find the axes (colour,direction, choice) are varying through time making it hard to tell what is orthogonal to what when plotted on axes that are constant in the figures.
2022-10-06 21:21:21 @StanDehaene @GabyMohamady @jcrwhittington When @timoflesch, @summerfieldlab , @saxelab reanalyse that data it looks almost perfectly factored. After a long discussion during Timo’s viva we decided this didn’t necessarily conflict the original analysis because of some subtleties in their pipeline.
2022-10-06 15:44:51 @summerfieldlab such a sweetie :)
2022-10-06 15:43:27 @akbirkhan @jcrwhittington neurons can't fire negative spikes :)
2022-10-06 13:55:55 @jcrwhittington @StanDehaene in a very similar task in mouse pfc, we also assumed these would be the obvious factors, but @GabyMohamady is finding a beautifully factorised rep, with a different (in fact better!!) set of factors. Mohamady is there a link to NM talk?
2022-10-06 13:50:27 @jcrwhittington @StanDehaene Another possibility is that the factors we assume as scientists are not the factors that the brain sees. I.e. these are not the factors: https://t.co/6bapNhSSXG
2022-10-06 09:37:53 Confirmed: Don't be afraid of equation 113 or these beautiful figures. The appendices contain a treasure-trove of insights and are written with a real sense of humour, and the reader in mind. https://t.co/EnCTtPqT1d https://t.co/rx9uClyUMJ
2022-10-06 09:33:17 RT @jcrwhittington: Appendices are a real delight to read - offering deep intuitions about grid cell behaviour - and the hand-drawn figures…
2022-10-06 09:32:23 RT @dorrell_will: New pre-print with Peter Latham, @behrenstimb &
2022-10-06 09:32:14 RT @jcrwhittington: Brilliant and beautiful work from the ultra-talented @dorrell_will! We provide a general formalism of neural representa…
2022-10-06 09:32:00 This is really pretty. The fantastic @dorrell_will and the scruffy @jcrwhittington argue neural representations should be "actionable": Internal representations can predict consequences of actions. They formalise this with group theory and predict a host of grid cells properties https://t.co/XSdAb62THv
2022-10-06 05:32:10 RT @SuryaGanguli: Our new paper https://t.co/gPIipr7Wag shows how two biologically plausible properties (non-negativity &
2022-10-06 05:31:34 RT @behrenstimb: From the amazing @jcrwhittington, with @dorrell_will, @SuryaGanguli, a mathematical justification for the neuron doctrine!…
2022-10-06 05:31:15 RT @jcrwhittington: If everything's a big manifold, why do neurons often code for human-interpretable factors? In https://t.co/xCaJbmtwwi w…
2022-10-06 00:06:34 Tomorrow’s tweeprint :) https://t.co/JV9adwTO3o
2022-10-05 23:44:14 From the amazing @jcrwhittington, with @dorrell_will, @SuryaGanguli, a mathematical justification for the neuron doctrine!! Explains why single neuron responses are often so beautiful in naturalistic (composable) tasks and shows how to get these disentangled reps in deep-nets. https://t.co/0fOcSwCek7
2022-10-04 20:59:30 RT @TFletcher: As dreams go, this is not up there with Martin Luther King.
2022-10-01 12:46:36 When hardening conkers in the oven, should I put the hole in the conker before of after heating it.
2022-09-30 21:36:21 This is amazing. @neuromatch team are my heroes. https://t.co/FesHjRhWFf
2022-09-29 15:32:13 RT @jcrwhittington: Review finally out! It goes through hippocampal representational models and what they tell us about building cognitive…
2022-09-29 15:32:09 A tome on all the recent hippocampal models from @jcrwhittington https://t.co/IaVxK3yCF2
2022-09-29 05:18:41 RT @behrenstimb: Final version of @nikasamborska's awesome generalisation paper now online: https://t.co/txQz0atK8eData: https://t.co/a6j…
2022-09-29 04:13:22 @dileeplearning Thanks Dileep :)
2022-09-28 21:27:37 @nikasamborska updating not installing :)
2022-09-28 21:27:07 RT @nikasamborska: mostly just happy it took Tim longer to write this tweet than to run my python code (including installing matplotlib) :)
2022-09-28 21:25:10 RT @behrenstimb: @hugospiers @nikasamborska @wellcometrust Watch out for AkamLab - it will rock!!
2022-09-28 21:25:02 @hugospiers @nikasamborska @wellcometrust Watch out for AkamLab - it will rock!!
2022-09-28 21:19:37 Original tweet storm (March 2021!!) here:https://t.co/P2r8UBCi3gSuper exciting for me, as first rodent paper to come out of the lab. Look out for more cool stuff to come :)
2022-09-28 21:19:36 Final version of @nikasamborska's awesome generalisation paper now online: https://t.co/txQz0atK8eData: https://t.co/a6jQdhKUYQCode: https://t.co/aHmoX2W1qFExcitingly: last author Thomas Akam has just won a @wellcometrust fellowship to start his own lab - look out for jobs!
2022-09-28 19:20:03 RT @WaltonLabNeuro: Many congrats to our lab head Mark (or we should say Prof Walton)!!
2022-09-28 13:47:33 In 10 mins! https://t.co/5FcOxCKXnO
2022-09-28 06:19:39 In 40 minutes!!! https://t.co/wOfINLTyMK
2022-09-28 04:40:52 RT @gautamcgoel: Just to give a data point: when I applied for postdocs, I was offered $70k/year at Berkeley and $96k/year at ETH. Honestly…
2022-09-27 15:26:51 @drghirlanda @GabyMohamady Nothing written yet (or for a few months). Can't you back-watch @neuromatch ?
2022-09-27 14:50:42 I know I am biased but I think this is the coolest thing in the history of the world. The neural representation completely reveals the algorithm the mice uses to solve a complex abstract inference, and store a long probabilistic sequence of actions. https://t.co/o4IDEei6hw
2022-09-27 14:47:58 RT @GabyMohamady: Want to know how mice do clever inferences the first time they see a new task? I’ll be presenting our latest findings wit…
2022-09-27 14:19:46 RT @neuromatch: Publishers may simply switch to charging excessively high fees to authors and their institutions, effectively pricing out a…
2022-09-27 14:19:43 RT @neuromatch: @WHOSTP recently committed to making all federally funded research free to read without embargo at publication. This is a h…
2022-09-27 14:16:24 RT @neuromatch: Scholars in all disciplines &
2022-09-26 21:42:24 RT @lastpositivist: Just got word that one of the best papers I've read in ages was rejected, after 8 months in review, on the basis of a r…
2022-09-26 15:46:18 @harootonian No - that would be sweet but unfortunately not - hence the exchange rate makes it hard to visit US
2022-09-23 10:57:38 @hugospiers What's more amazing is the bit that is *not* accounted for by group dynamics. They have individual relationships so that little guy knows his mate is in the wrong group. Wouldn't get that in a fish shoal :)
2022-09-23 07:55:07 Philomena Cunk is on every grant panel. https://t.co/PwHwjSZST0
2022-09-22 18:13:03 @demishassabis Congrats Demis. Amazing stuff. And very good planning to wait for the best exchange rate for 40 years :)
2022-09-21 19:58:41 RT @SWC_Neuro: Applications now open! Join the SWC PhD Programme World-class training in theoretical and systems neuroscience Fully-fu…
2022-09-21 09:42:29 @PhilCorlett1 This would make paper writing great fun. Who can I make retire today?
2022-09-20 18:19:59 Transparency and accountability. Treasury refuses to publish UK economic forecast https://t.co/H322hIzw1h
2022-09-20 17:44:18 RT @TPVogels: Applications for the @isiCNI Imbizo'23 are open now. Apply before Oct. 20th to join us from Apr 13 to May 7th, 2023 to learn…
2022-09-20 13:38:29 Amazing https://t.co/FQYtbsXhn9
2022-09-20 13:38:24 RT @AdamCSharp: I celebrated officially getting my PhD last week by buying myself some things online and when they arrived saying to the de…
2022-09-20 11:04:29 @xandvt @BBCMorningLive Congrats Xand - awesome news :)
2022-09-19 15:48:56 @MelMitchell1 @criticalneuro @Tim_Dettmers @WiringTheBrain @blamlab @tyrell_turing @jcrwhittington Yeah of course. Sorry if I seemed down. It wasn’t about your question. I was just scrolling through the thread and it all just seemed so random
2022-09-19 14:27:42 @criticalneuro @Tim_Dettmers @MelMitchell1 @WiringTheBrain @blamlab @tyrell_turing @jcrwhittington Yeah maybe. This is the bit that he shouldn't have written: https://t.co/SZM9rolYry
2022-09-19 14:23:38 @criticalneuro @Tim_Dettmers @MelMitchell1 @WiringTheBrain @blamlab @tyrell_turing @jcrwhittington Here the hype is not from the journalists. It's from the academic commentators, many of whom I suspect haven't read the article nevermind the papers.
2022-09-19 14:22:51 @criticalneuro @Tim_Dettmers @MelMitchell1 @WiringTheBrain @blamlab @tyrell_turing @jcrwhittington ... of explaining that what has been shown is a mathematical relationship between models. Obviously, it starts with a bit of fluff, but the main article basically gets the spirit right. Then twitter takes it to all sorts of weird places.
2022-09-19 14:20:34 @criticalneuro @Tim_Dettmers @MelMitchell1 @WiringTheBrain @blamlab @tyrell_turing @jcrwhittington This whole thread is kind of an education for me. There are mathematical equivalences shown in Ramsauer and Hochreiter, in Krotov and Hopfield, and in our paper. The article does a reasonable job (not perfect but not totally out there) ...
2022-09-19 04:45:11 @CharanRanganath @MelMitchell1 @WiringTheBrain @blamlab @criticalneuro @tyrell_turing Hi Charan. My thoughts on this matter are limited to a mathematical relationship between the Tolman Eichenbaum Machine and a (modified) transformer. The article actually explains this pretty well if you ignore the fluff at the start.
2022-09-18 20:23:18 @MelMitchell1 @WiringTheBrain @blamlab @criticalneuro @tyrell_turing I don’t think any of the scientists involved would translate that rather precise statement into the meaningless statement “the hippocampus is a transformer”. James and I certainly wouldn’t.
2022-09-18 20:21:22 @MelMitchell1 @WiringTheBrain @blamlab @criticalneuro @tyrell_turing It’s obviously a silly statement. But most of this article is pretty good and stays on firm ground. What is true is that there are close mathematical parallels between transformers and current models of Hippocampus. That’s what the research papers show.
2022-09-16 23:13:56 @brody_lab @dikshagup @PrincetonNeuro Woohoo congrats @dikshagup
2022-09-16 13:18:30 @potentinocastle Yes! And 3rd from right :)
2022-09-16 08:22:25 @LHuntNeuro I accept mine are both the same, but the subject matter is definitely more on point. https://t.co/210BOY3C6r
2022-09-16 08:02:31 @LHuntNeuro Interesting. That genuinely hadn't occurred to me. Maybe that will be how to amplify the crisis at 50.
2022-09-16 07:42:48 My mid-life crisis is a gradual rejection of monochrome socks. Progress to date. https://t.co/7XckJbxplO
2022-09-16 00:46:52 RT @StephensonJones: First preprint from the lab!! All credit goes to @cesca_gst &
2022-09-12 10:36:07 RT @Haggis_UK: This sort of announcement could only be made in the UK. https://t.co/MeU1YNDZT0
2022-09-10 16:18:05 RT @Tattooed_DM: No one let @trussliz near Sir David Attenborough.
2022-09-09 18:00:47 A really stunning set of ideas with this being a real highlight. https://t.co/X3Ozrvm2k5 https://t.co/j2grnwXGSb
2022-09-09 16:29:16 @SussilloDavid Almost never successful unless the editor has given a clear reason. If they have then the same three rules apply.
2022-09-09 12:51:46 If you are writing appeal letters, I think that broadly there are only 3 successful successful arguments:1. The reviewer is factually wrong2. The reviewer has not understood3. You have new data proving the reviewer is wrong. "We disagree with the reviewer" never works.
2022-09-08 18:07:31 RT @wellcometrust: Calling all neuroscientists. Did you know that you're eligible for our Discovery Research funding schemes? [1/11]…
2022-09-06 21:40:30 65% left. 31% right https://t.co/vp6AHDaqx9
2022-09-05 17:03:05 @cleclouze Pretty much the same in the U.K. Thatcher and Blair
2022-09-05 16:29:20 @the_mindwanders I don’t think so, but one amazing thing ->
2022-09-05 16:02:32 My 6yo daughter is on her 4th prime minister.
2022-09-04 09:27:08 RT @neuralreckoning: Last day is tomorrow!
2022-09-01 21:34:09 “What appears to be intelligence … may in fact be a mirror that reflects the intelligence of the interviewer”This is what I’m aiming for when my group ask me about their projects. https://t.co/nxLmEQEgto
2022-08-30 21:16:31 @yeewhye @HeathrowAirport It’s totally insane. We queues for an hour last week, surrounded by toddlers and babies. Everyone else was through in 5 mins.
2022-08-30 11:46:04 This looks like a great opportunity and the job ad signals a cool new direction in Wellcome funding (->
2022-08-30 11:40:53 RT @MTC_Brown1: Wellcome Job Alert! Come work with me and a brilliant Data for Science and Health team to shape and fund mental health…
2022-08-29 17:48:13 @drkjjeffery Congrats Kate!!!!
2022-08-29 13:51:23 @gershbrain Yeah whether to pay reviewers is complicated. I would do it, but I understand the risks. For me, editors is a different thing. It is a lot of time. It is a job. I think people should be paid for their work.
2022-08-29 13:48:38 @gershbrain It's the same issue as "Free internships". If you don't pay people, you only get certain people. And they benefit.
2022-08-29 13:46:52 @gershbrain When you are running a big journal, your editors go awol. Who is keeping track of all the papers so that authors don't wait 2 years. You need actual staff to manage 1000 papers a month (or more). etc. etc. etc.
2022-08-29 13:44:40 @gershbrain New technologies emerge. Who is going to make the new back end? What if you want your journal references properly cross-refed? How are you going to pay the cross-ref fees?
2022-08-29 13:43:36 @gershbrain On top of this, big journals have costs that small journals don't have. New policies are developed and debated. Ethical problems emerge. You can be EIC of a small journal 10% of your time. Try being EIC of Nature 10% of your time.
2022-08-29 13:41:11 @gershbrain Sure but this way you are paid $100 per hour to review and edit papers. Profs in many countries are paid 1/5th that for the same job. This social contract is why most editors are white men from the US east coast. It promotes a workforce who can afford to work for free.
2022-08-28 21:04:29 @GabyMohamady @smotus :)
2022-08-28 19:05:54 @Neurocircuits @NikkiPelot @blsabatini @MarinaP63 I’m not aware of any specific policy on this, but if methods are identical I personally think it’s fine to lift descriptive text if you are clear that is what you are doing.
2022-08-28 18:03:54 @gershbrain @hholdenthorp I don’t buy that publishing is or should be free. Your solution just hides the cost in Harvard salaries. The right solution is surely to be honest about the cost and fund it centrally (no APCs. Funding agencies directly funding journals).
2022-08-27 17:22:08 Just parched. https://t.co/IIxqvv5QIY
2022-08-27 17:20:06 Southern England could do with a drink. https://t.co/zK05oi97Ro
2022-08-27 07:13:35 RT @mbeisen: You know what I don't want to ever hear again? A single complaint from anyone about how efforts to fix science publishing are…
2022-08-25 19:20:02 !!!! https://t.co/5TJ5wyH0n1
2022-08-25 05:09:33 RT @cian_neuro: UK academics’ pay wasn’t even keeping up with the v low inflation of the past decade, never mind the crazy rates of today.…
2022-08-24 17:10:55 @SWC_Neuro @trabranco @royalsociety Woohoo. Awesome news :)
2022-08-24 17:10:29 RT @SWC_Neuro: Congratulations to SWC Group Leader Professor Tiago Branco (@trabranco) on being awarded the Francis Crick Medal and Lecture…
2022-08-20 07:43:09 RT @MarinaHyde: “Liz Truss is a planet-sized mass of overconfidence and ambition teetering upon a pinhead of a political brain.” https://t.…
2022-08-19 11:21:56 RT @HelenCBarron: 3rd post we are recruiting now live! Postdoctoral Neuroscientist for exciting cross-species project, focusing on neural m…
2022-08-16 10:51:58 This is an absolutely delightful piece of work. You can see Ethan talk about it at 1:39 here https://t.co/eJ5eR780ry. It is of the most engaging 10 minute talks, by one of the most joyful scientists, you will ever have the pleasure of hearing. https://t.co/cT1gvgBPU6
2022-08-15 22:31:51 RT @HelenCBarron: Second post we're recruiting: Postdoctoral Neuroscientist, joint with Jill O'Reilly for some exciting work on neural comp…
2022-08-13 12:01:15 https://t.co/43hPWkfYDS
2022-08-13 11:35:56 Learning comp neuro is tough in Woods Hole (Thanks @Nancy_Kanwisher for this special treat). https://t.co/NzBqHFpMQn
2022-08-07 17:36:57 @ElDuvelle :) thanks El. Not this time but my sister lives here so I’ll be back ….
2022-08-06 13:12:41 @alfairhall @gershbrain Currently in New Hampshire mountains but going to Woods Hole (via Boston) next week so may be able to take you both up on kind offers!!
2022-08-05 08:38:11 LHR ->
2022-08-04 09:26:12 RT @brian_bilston: Here’s a short poem about the #ToryLeadershipContest. It’s called ‘Outright Winner’. https://t.co/z7JchWBpm9
2022-08-03 08:11:14 @TPVogels I agree with the second part.
2022-07-31 17:15:46 @TonyZador And when I specify that the backhand is left-handed, not the player, it makes no difference. "a photograph of a frog hitting a left-handed double-handed tennis backhand, digital art"Experiment over. I resign. https://t.co/YMJYs0k40M
2022-07-31 17:14:29 @TonyZador yep - when I ask for a double-handed backhand it gets very confused: https://t.co/QDxfhC4naQ
2022-07-31 16:56:43 @TonyZador Huh - you even have two backhands. Here are my 4. Weird indeed ... https://t.co/nFxct5aRTu
2022-07-31 15:21:03 A realistic photograph of a half-frog-half-girl drawing a picture of a cat on an easel (viewed from the side)(Not sure why she is drawing on the side of the picture or why the cat is also drawing). https://t.co/U55wGBF8Xf
2022-07-31 15:21:02 "A photograph of a left-handed frog hitting a tennis backhand, digital art"All the frogs were right-handed mostly hitting forehands. https://t.co/0Iqge3LxyD
2022-07-31 15:21:00 But "A photograph of a frog painting a picture of a cat"(all the images were this way round - not sure why cats are more likely artists than frogs). https://t.co/WTof82Vxa5
2022-07-31 15:20:59 For example"A realistic photograph of a half-frog-half-child playing football at Wembley wearing an England strip" https://t.co/pwCfGxjSW7
2022-07-31 15:20:58 I got Dalle-access. It's awesome fun. I learnt 2 things so far. 1) kids have strange imaginations. 2) It can be absolutely amazing at understanding language, but it often makes mistakes understanding basic scene composition
2022-07-29 04:18:24 RT @SaxeLab: Job alert! Postdoc in deep learning theory @GatsbyUCL We're currently interested in theories of systematic generalization, a…
2022-07-28 05:40:02 This is totally awesome from Ivan Voitov and @TFlogel https://t.co/6f9VzUSvjN https://t.co/MXlzf0pexG
2022-07-25 16:12:05 RT @gershbrain: @naoshigeuchida are still recruiting a postdoc to do theory-driven experiments on learning and decision making in rodents.…
2022-07-25 11:26:56 @TPVogels Does this mean you're sending it to eLife now?
2022-07-19 06:55:01 RT @SaadJbabdi: We’re hiring! If you’d like to come work in a fun lab and a beautiful city, come and join us. See details of the post…
2022-07-18 05:21:49 RT @ColginLab: @Ananya_MC Please add your name and associated information to this document, which a lot of editors use for finding reviewer…
2022-07-17 18:01:23 @neuralreckoning Eating ice cream never makes anything worse in any circumstances.
2022-07-16 17:47:56 This is going to be an absolute beast on “Guess the mystery object - Antiques Roadshow 2122.” https://t.co/euFR5NSw16
2022-07-14 18:12:31 @aluskin @anne_churchland Oh cool - How do we know this is direct rather than via EC or reuniens?
2022-07-13 06:45:44 @micahgallen @K_M_M_Walker @jcrwhittington @GabyMohamady @AnnaShpektor Thanks both. It was lovely to spot friendly faces in the audience :)
2022-07-12 14:05:10 @BerndPorr Could definitely be about the locations paths/goals in different tasks. But can’t be simply that because of the coherent rotation. Also must count the ABCD
2022-07-12 14:03:56 @BerndPorr We’re not sure yet what causes the rotation between tasks but we have a few hypotheses to work with. Will keep you posted as we dig further into the data!
2022-07-12 11:11:20 After spending 2 hours dicking around converting all my keynote fonts and transitions to powerpoint, I am now ready and raring to go :) https://t.co/slU072CSml
2022-07-12 08:09:11 @micahgallen @jcrwhittington @GabyMohamady @AnnaShpektor This talk is definitely a mixed bag. It has theory, rodent behaviour, rodent cells, human cells, human fmri, human MEG. Hopefully people can see the links :)
2022-07-12 06:14:17 @criticalneuro @jcrwhittington @GabyMohamady @AnnaShpektor Thanks Ida - No preprints on these studies yet (actually both studies will collect more data before we publish so a while off).
2022-07-12 06:03:00 If you’re at FENS today I’m giving a talk at 1.30 hall C. Will briefly cover Tolman Eichenbaum machine (@jcrwhittington) and present 2 new datasets. This one from @GabyMohamady and a human one from @AnnaShpektor, mixed in with nods to a few previous lab highlights. https://t.co/7s3qCvs5w4
2022-07-09 14:46:51 RT @GabyMohamady: Looking forward to a double bill of posters at #FENS2022!Come watch videos of mice performing zero shot inferences in ta…
2022-07-09 07:14:37 @jjfahrenfort Easier on computer than phone :)
2022-07-08 12:56:29 RT @fmrib_karla: @behrenstimb this is the greatest thing you have ever caused to happen Behrens. time to retire.
2022-07-08 11:37:14 RT @nikasamborska: if you are curious what working with @behrenstimb is actually like @BakermansJJW perfectly summarised it here
2022-07-07 21:00:41 Top fact: Jacob made a supervisor simulator game where you get to be me. You have to avoid the students and emails and collect the whisky. Play it here:https://t.co/dsRuU8Gdsm https://t.co/BucLBgBrFG
2022-07-07 17:17:54 Dr @BakermansJJW !!!! https://t.co/vUNKWg5oWW
2022-07-07 17:00:41 RT @HackedOffHugh: Dear World, You may be wondering what happens next in terms of the British constitution. The answer is that 3 newspaper…
2022-07-06 21:03:24 RT @GeorginaMacKen4: Wellcome relies on experts from around the world to help us decide which application to fundWe are now seeking expre…
2022-07-06 21:03:21 This is a great way to get a broader spectrum of people making the important decisions. https://t.co/BbW1cVV325
2022-07-06 09:30:58 @hugospiers Was this the problem? https://t.co/1OIIhKovQM
2022-07-04 16:21:34 @TobiasUHauser @FENSorg @FensKavliNet Pretty sure the ERC isn't the problem! Can we invite Macron?
2022-07-04 16:21:02 RT @TobiasUHauser: @behrenstimb Come to our @FENSorg @FensKavliNet panel session and ask the ERC vice president directly... Would be curiou…
2022-07-04 07:33:02 Hey Europe - can you let us into Horizon please? I realise we are behaving like spoilt children elsewhere, but we are not responsible enough to babysit ourselves. https://t.co/iMjDp78Yc5
2022-07-03 09:26:54 @YardenJCohen @StokesNeuro Thanks man. Who knows what this wave will bring https://t.co/LMhcMdWt2S
2022-07-02 21:50:03 @nikasamborska Jesus. Hope she’s ok :(
2022-07-02 17:56:37 @TimoWitten @StokesNeuro Pretty sure the lab is more effective without me involved.
2022-07-02 17:42:44 @AthenaAkrami @StokesNeuro Yep. Seems like most of London has it right now though. Stay safe!
2022-07-02 16:54:26 @StokesNeuro Looks like a really fun schedule an awesome evening to be out in Oxford
2022-07-02 16:53:02 @StokesNeuro Thanks man. Have a really fantastic day.
2022-07-02 16:11:07 @bradpwyble @StokesNeuro So far it has got easier each time. Don't know about this one yet. Quite a while since my last shot.
2022-07-02 15:23:02 @AdamHantman @StokesNeuro @ONS ... sea shells on the sea shore? Dunno - there is no T-cell test in the survey I'm on. Guess I could pay privately, but unlikely to :)
2022-07-02 15:16:54 @AdamHantman @StokesNeuro I'm part of the @ONS survey so they check my antibodies every month. They keep telling me I have lots of them. Don't seem to be very good ones though.
2022-07-02 15:15:45 @criticalneuro @StokesNeuro They've all been about 6 months apart. I have basically caught it every time there is a wave in UK (except Delta). Caught it from the kids this time.
2022-07-02 15:03:47 4th time. Totally gutted to miss @StokesNeuro celebration in ox today. https://t.co/DpV8ogDpS4
2022-07-02 08:15:09 RT @biorxiv_neursci: Noradrenergic signaling in the rodent orbitofrontal cortex is required to update goal-directed actions. https://t.co/…
2022-07-01 09:25:13 Quite impressed with the nominative determinism in yesterday's Tory sex scandals. BJ and Pincher fulfilling their predestined roles.
2022-07-01 06:05:44 @ErieBoorman Cool bananas
2022-06-30 19:16:26 RT @GatsbyUCL: Congratulations to Joint Gatsby-@SWC_Neuro Group Leader Andrew Saxe (@SaxeLab) on the Schmidt Science Polymaths Award! Andre…
2022-06-30 19:16:15 RT @Dee_Kay_Jay: On my way home from serving on the @OxfordWIN Public Engagement external advisory board. An amazing amount of activity…
2022-06-29 15:56:01 RT @MTC_Brown1: A privilege to work with @KirstyDonald7, @mahmoudbukar and a brilliant steering group digging into current areas of distinc…
2022-06-28 13:47:59 RT @abhii_mit: Really sad to hear about the passing away of Colin Blakemore. Colin was the reason I became interested in neuroscience, afte…
2022-06-27 22:06:03 RT @simonoxfphys: Did you know that a UK PhD stipend is so little that it puts recipients below the absolute poverty line?UK PhD stipend…
2022-06-27 05:51:42 RT @ProfBrianCox: The reason the UK will have the lowest growth in the G7 next year is Brexit. We're not going to reverse the decline until…
2022-06-24 19:32:02 @demishassabis Thanks Demis. Now scientists of the future will be able to flip past our page whilst they’re looking for yours :)
2022-06-24 17:36:59 @JonAMichaels @EricMTrautmann Thanks Jon! Yes here is the context in full:https://t.co/1gXIGrQc5y
2022-06-24 15:26:05 @criticalneuro definitely a surreal experience :)
2022-06-24 10:30:18 RT @SeanNolan_: The break up was described as cordial
2022-06-24 06:25:03 RT @JolyonMaugham: *or* keep going, surely?
2022-06-23 21:02:26 @erinkendallb Isn’t illegible penmanship a prerequisite to be a scientist?
2022-06-23 18:20:03 @Dick_retired Yes it must have been scary. Fortunately they since moved to a fountain pen, making it a bit easier for the newbies :)
2022-06-23 15:44:10 @caswellcaswell @smfleming The underlining was totally spontaneous. I felt like such a rebel.
2022-06-23 14:34:27 @caswellcaswell @smfleming Yes you practise first, but about an hour beforehand with no nerves. My practise attempt was a work of art, but when I got there I made the horrid scribble you see above.
2022-06-23 13:39:55 @potentinocastle scary :)
2022-06-23 12:44:31 @smfleming Thankfully, they're paying close attention to what you're up to so no danger of that. Was lovely to be there with Cathy. We skulked off together a couple of times to escape the pressure cooker :)
2022-06-23 12:14:58 RT @AnnaShpektor: Not sure Tiwtters magic will be enough to solve this one but any chance you know a Russian speaking kids psychologist bas…
2022-06-23 09:24:37 @martopix Yes the people present were almost as awe-inspiring as the people past. 5 Nobel laureates on that page. 3 women (Donna Strickland, Frances Arnold, Ada Yonath) and 2 men (Queloz and Ben Feringa).
2022-06-23 08:58:16 Cathy Price and I signed this book yesterday next to heroes of science past and present. I felt like a charlatan and a fan-boy. So nervous I could barely write my name, then I nearly fell off the stage. Parents and Mrs B v proud though. https://t.co/q2a5dfGLxv
2022-06-23 04:35:36 RT @andpru: definitely
2022-06-21 16:46:55 Portrait of @sjblakemore at the @royalsociety exactly where it should be. https://t.co/Oh9AVQUKOk
2022-06-21 16:39:33 Peer review written by Charles Darwin for the royal society. “… Not precise enough owing, no doubt, to the author’s attention being directed to other objects …” https://t.co/X1FxDkQvjk
2022-06-21 09:17:41 RT @hugorifkind: Mad how we are still in this world where, out of fear and political triangulation, nobody can quite bear to admit how bad…
2022-06-21 09:04:44 RT @hugorifkind: Remember how they used to ridicule predictions like this? You wonder how they sleep.
2022-06-21 06:06:59 RT @BoyNamedShit: God this is such a good prompt https://t.co/lhVpNtU4BG
2022-06-20 21:44:00 RT @AnnaShpektor: Anyone has a spare room for 3 nights somewhere near Den Haag or Utrecht for a Ukrainian family with 2 cats? They had to w…
2022-06-20 21:40:28 @nikasamborska I’ll ask her tomorrow:)
2022-06-20 06:44:54 @nikasamborska I thought so too. Twitter seems unimpressed…
2022-06-19 20:18:42 “Picatto” https://t.co/H7F4fLfErU
2022-06-17 09:06:52 @Dick_retired @biorxiv_neursci I totally missed this tweet. Thanks Dick! That is a really nice thing to say :) I also got a lovely email from @lynn_nadel, who has recently proposed sthg like this should happen https://t.co/yTvqw89VNzand a few pointers to related effects outside of the naturalistic context.
2022-06-17 06:20:36 RT @kaarinaho: Very excited that our paper 'System alignment supports cross-domain learning and zero-shot generalisation' is out now in Cog…
2022-06-16 17:58:38 @Hui_Hui_Zhang @AvitalHahamy @jmourabarbosa Awesome - thanks for pointing us to this (and thanks @luohuan_huan). Nice to see a relationship between task activity and what happens in naturalistic movie watching. We'll be sure to cite in revision :)
2022-06-16 07:41:50 RT @Ishwariya13: I remember breaking down to my Ph.D. advisor about how stupid I felt while troubleshooting a problem in my project when sh…
2022-06-16 06:53:50 RT @mikegalsworthy: I never liked @AndrewMarr9 before...But this - THIS - on the government’s Rwanda policy is absolutely on fire - and I…
2022-06-16 05:47:32 RT @MRCBNDU: Many congratulations to our very own @HelenCBarron on being awarded @UKRI_News Future Leaders Fellowship. @MRCBNDU ’s newest G…
2022-06-15 20:04:24 RT @SonjaBHofer: Mitra Javadzadeh's study on causal inter-areal communication is now out in @NeuroCellPress with new data! We find that com…
2022-06-15 15:15:21 @jmourabarbosa @AvitalHahamy Thanks Joao - will take a look :)
2022-06-15 15:14:27 Never has anything been more deserved in the history of humanity. https://t.co/ZEhHPCG1bz
2022-06-15 08:34:02 RT @frankieboyle: The Rwanda flight on the anniversary of Grenfell says everything about the direction we’ve traveled in the last five years
2022-06-14 17:54:48 New paper with the awesome @AvitalHahamy showing reactivations of past events whilst understanding a movie/audio-book. The brain reactivates past events at event boundaries, linking relevant events across time. https://t.co/PXC3aD15NS
2022-06-14 06:12:03 RT @CeliaHeyes: Five-year independent postdoc opportunities at All Souls College, Oxford, including 2 in life or environmental sciences (e.…
2022-06-11 06:52:30 @lessc0de Pretty unambiguous with the ERG.
2022-06-11 06:51:52 RT @NeilBurgess10: @behrenstimb Indeed - here’s who he’s literally owned by https://t.co/blr03NP21o
2022-06-11 06:14:29 @TrackingActions @JMontijn @_rdgao Hey @elife has a pool of early career reviewers. I suspect you got added (can check on Monday). But reviewer selection is still in the hands of individual editors, and not all use the ecr pool. The reviewer invitation problem is on next year’s list to tackle.
2022-06-11 05:25:44 Johnson now 100% owned by hard right. Many disasters here but one is presumably that agreement over U.K. partnership with ERC is now even further away. https://t.co/RHdfHmbET0
2022-06-10 10:10:34 RT @SonjaBHofer: Fantastic lab in amazing @SWC_Neuro research environment. This is a great opportunity!
2022-06-10 10:10:27 RT @AnnDuan2: Our lab is hiring a Research Assistant, please RT!Come join our team to study the neural circuit mechanisms underlying flexi…
2022-06-09 11:15:45 RT @SagivYotam: Have you ever been at the grocery store and thought “I wonder how I should prioritize replay in the absence of rewards”? ……
2022-06-09 05:52:50 @neuralreckoning @gershbrain @ElDuvelle @stevenmweisberg @danielmabuse @StePalminteri Who are the employers here? They are just us in different roles. Presumably you are also on these panels, Dan. I think what Sam is suggesting is that collective action we can take in *these roles* (where we actually have power) would be more impactful.
2022-06-08 18:39:02 RT @Alex2Thiele: Non human primate Neuroscience PI job https://t.co/0usmRQzmth
2022-06-08 12:58:05 @guyi That doesn't help. I still have to scroll through a million windows to find it :)
2022-06-08 12:51:36 @KiaNobre Congrats Kia!
2022-06-08 10:24:00 @fmrib_steve Think how much easier it would be to get grants for "pure neuroscience" than "basic neuroscience" though. I would never have to bother you again.
2022-06-08 09:20:40 @neuralreckoning This looks awesome!
2022-06-08 08:35:06 Jupyter notebook people. How do you stop your notebook being just another tab/window? Since I started trying to build one, I have spent 50% of my time looking for the right tab.
2022-06-07 17:55:28 Yes! So much better than “basic”. “Pure neuroscience” https://t.co/ZZLfHh65gX
2022-06-07 05:18:11 @Matt_Nour @mbeisen Awesome - thanks Matt :)
2022-06-06 19:55:55 @crewilson @OxfordWIN :)
2022-06-06 19:29:54 .@OxfordWIN recently held a confidence vote in our leader. We voted 97% in favour. #heidijb4pm
2022-06-06 18:22:52 @mbeisen @meganinlisbon :)
2022-06-06 17:26:04 @meganinlisbon This is a true reflection of how buggy my coding is.
2022-06-06 17:00:53 @criticalneuro @AthenaAkrami Hope it's a mild one. Get some rest :)
2022-06-06 16:52:15 whoops - obviously important + 0.5* (useful-important):)
2022-06-06 16:42:59 if anyone has a trained word embedding to hand, can you tell me what is nearest to 1/2 way between "Important" and "Useful". i.e. what is important + 0.5* (important-useful). (also feel free to guess without the word embedding). @mbeisen
2022-06-06 07:12:17 The best letter yet, nailing the key points. Reminder that there is a heart somewhere in the Tory party. Hope it keeps beating. https://t.co/sLgbn9LVW5
2022-06-03 05:51:21 RT @Life_Disrupted: Tobias Ellwood has crossed a red line. Tories are not allowed to tell the truth, that Brexit is a disaster and the only…
2022-06-02 06:25:16 RT @fredmstoll: New Paper alert in @eLife Inhibitory control of frontal metastability sets the temporal signature of cognitionFrom exc…
2022-06-01 19:06:06 @Raamana_ @neuralreckoning @sciencedave_wv @KordingLab @WorldImagining @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery etc etc. Note that there is a conflict between wanting this work done (it is an enormous amount of work) and wanting low APCs.
2022-06-01 19:05:16 @Raamana_ @neuralreckoning @sciencedave_wv @KordingLab @WorldImagining @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery Diversity of our editorial boardhttps://t.co/jq2DrA59LQ
2022-06-01 19:03:22 @Raamana_ @neuralreckoning @sciencedave_wv @KordingLab @WorldImagining @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery Some more recent stuff on gender imbalanceshttps://t.co/nvllAfp26I
2022-06-01 19:01:47 @Raamana_ @neuralreckoning @sciencedave_wv @KordingLab @WorldImagining @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery I'm sorry I missed this tweet. We profess commitment to open science and we publish this stuff. Here is an example with many of the stats you are asking for:https://t.co/aDgnRLB9hz https://t.co/KjG75hOzVA
2022-06-01 18:20:12 @Raamana_ @WorldImagining @neuralreckoning @sciencedave_wv @KordingLab @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery And some Asian govts pay money when you publish in big journals and we are not on those lists.
2022-06-01 18:19:01 @Raamana_ @WorldImagining @neuralreckoning @sciencedave_wv @KordingLab @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery Oh right. If you mean we should run marketing drives in Asia to increase awareness then I totally agree. Although it will be tough. A lot of our reputation initially stemmed from HHMI/Wellcome/Max Planck which are not big in Asia
2022-06-01 18:13:48 @Raamana_ @WorldImagining @neuralreckoning @sciencedave_wv @KordingLab @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery The question then is what to do. Our response was to run a recruitment drive for Asian editors as we thought this might have a positive effect on both potential problems.
2022-06-01 18:12:09 @Raamana_ @WorldImagining @neuralreckoning @sciencedave_wv @KordingLab @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery I think you’re missing the point here man. Those numbers will look bad. We don’t know the reason. One potential reason is we are tougher on Asian authors. Another is Asian authors try more journals before us on average. We know the second is true.
2022-06-01 17:25:17 @neuralreckoning @Raamana_ @sciencedave_wv @KordingLab @WorldImagining @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery 2/2 However, eLife also has a lower reputation in Asia than Europe/NA so Asian authors are much less likely to send us their best work (we know this from other data). Impossible to tell whether the triage rate is biased or accurate.
2022-06-01 17:23:19 @neuralreckoning @Raamana_ @sciencedave_wv @KordingLab @WorldImagining @bradpwyble @GunnarBlohm @ElDuvelle @drkjjeffery Totally impossible to disentangle. eg eLife has a (quite substantially) higher desk-reject rate for asian papers than North American/European. This may be bias, and indeed it has provoked us to recruit asian BREs heavily. 1/2
2022-06-01 11:05:50 RT @OdedRechavi: Regarding the "peer review improves papers" argument - OK, but also writing the paper the way you want to write it improv…
2022-06-01 09:24:29 RT @feathered_ape: .@birchlse has written an open letter to @Cambridge_Uni to reconsider closing @nickyclayton22 Corvid Lab. Read the lette…
2022-06-01 09:24:26 RT @ElDuvelle: Sometimes, bad things happen and you can’t do anything… in this case, we can do something:Sign this open letter in support…
2022-06-01 05:33:32 @WorldImagining @bradpwyble @Raamana_ @neuralreckoning @KordingLab @GunnarBlohm @ElDuvelle @drkjjeffery More broadly though. If we are vilifying people for acting in their own interests we are nuts. Instead we should try to figure out a system that more closely aligns individual and common interests.
2022-06-01 05:30:05 @WorldImagining @bradpwyble @Raamana_ @neuralreckoning @KordingLab @GunnarBlohm @ElDuvelle @drkjjeffery But I would say even people in this thread are acting in our own interests. Personally I am fed up moulding my science to templates set by the whims of past Nature editors, being scared of reviewers in temporary positions of power in case I jeapordise a trainee’s career. Etc
2022-06-01 05:13:27 @WorldImagining @bradpwyble @Raamana_ @neuralreckoning @KordingLab @GunnarBlohm @ElDuvelle @drkjjeffery Here’s another one https://t.co/mCcReWfZAL
2022-06-01 05:11:27 RT @leslievosshall: @OdedRechavi @IgorUlitsky Yes true - we care about the discovery not journal name. And we support @eLife which is point…
2022-05-31 18:14:00 @emrobSci Yes it’s bonkers!
2022-05-31 04:43:22 @neuralreckoning @KordingLab @bradpwyble @brembs @prokraustinator @GunnarBlohm @ElDuvelle @WorldImagining @atypical_me @Raamana_ @drkjjeffery I have absolutely no clue what these entire mad threads are on about. Enjoying nevertheless:)
2022-05-31 04:41:31 And U.K. unis were a top destination for the brightest grads in Europe. Now there is no funding for them. Make-up of grad population totally changed. https://t.co/yhUE2grdAl
2022-05-30 19:48:14 RT @CathAMClarke: BREAKING: Johnson to bring back Old English letterforms for the Platinum Jubilee! The PM said: Many of þese letters Ƿere…
2022-05-30 19:02:39 @_stah @CousinAmygdala @eLife But seriously … I think this is a rate that will be attractive even to senior folk, who may nevertheless be doing some of it out of favour, and will be fair pay for the expertise and time that most (more junior) reviewers bring.
2022-05-30 18:58:50 @_stah @CousinAmygdala @eLife Jesus Christ US folk have a warped view of academia.
2022-05-30 18:52:36 @_stah @CousinAmygdala @eLife For me it’s not about compliance. It’s about paying for work. Simple as that. Running an entire industry on guilt and favours is asking for trouble. In my view. I realise others have good reasons to disagree though :)
2022-05-30 18:37:46 @CousinAmygdala @eLife https://t.co/B2JUK2lPtj
2022-05-30 16:28:06 @andpru But don't worry. (1) I think there are bigger problems to solve first. And (2) other people on the leadership don't agree with me, so I don't think it will happen in the near future at least :)
2022-05-30 16:21:19 @andpru Yes. $300
2022-05-30 14:02:52 @andpru @TrackingActions We have never paid reviewers. We pay BREs and SEs but nothing like as much as we should.
2022-05-30 13:59:13 @andpru https://t.co/sRS7pNO9e4
2022-05-30 13:45:44 @andpru How much should we pay?
2022-05-30 12:49:53 @neuralreckoning @WorldImagining @bradpwyble @Raamana_ @KordingLab @GunnarBlohm @ElDuvelle @drkjjeffery In our data there is definitely not a general age or gender effect on a simple binary measure of "support in principle"
2022-05-30 12:49:05 @neuralreckoning @WorldImagining @bradpwyble @Raamana_ @KordingLab @GunnarBlohm @ElDuvelle @drkjjeffery Similarly, I agree with you that there are many male senior people who are aggressively anti-change, but I have been encouraged by how many senior people of both genders have been extremely supportive of widespread change.
2022-05-30 12:48:01 @neuralreckoning @WorldImagining @bradpwyble @Raamana_ @KordingLab @GunnarBlohm @ElDuvelle @drkjjeffery ... of both genders. But there are also many ECRs who worry about how they will get jobs in the absence of the current system. They also worry about a massive shake-up of the system just when they need it to work for them.
2022-05-30 12:46:35 @neuralreckoning @WorldImagining @bradpwyble @Raamana_ @KordingLab @GunnarBlohm @ElDuvelle @drkjjeffery I wish it were as simple as this. We have tried polling this kind of question and I think it is a lot more mixed than you imply here. We have some imperfect stats on this that I will be able to share with you in the future. There are definitely many supportive ECRs ...
2022-05-30 05:45:34 @GunnarBlohm @neuralreckoning @atypical_me @Raamana_ @KordingLab @ElDuvelle @WorldImagining @drkjjeffery @CogCompNeuro @meganakpeters Def don’t worry at all. I am really pleased to have done it.
2022-05-30 05:41:53 @meganakpeters @GunnarBlohm @neuralreckoning @atypical_me @Raamana_ @KordingLab @ElDuvelle @WorldImagining @drkjjeffery @CogCompNeuro Cool. Awesome that this works so well. It’s a really fantastic idea.
2022-05-29 21:01:23 @GunnarBlohm @neuralreckoning @atypical_me @Raamana_ @KordingLab @ElDuvelle @WorldImagining @drkjjeffery @CogCompNeuro Obviously very pleased to have been ambushed in retrospect :)
2022-05-29 21:00:38 @GunnarBlohm @neuralreckoning @atypical_me @Raamana_ @KordingLab @ElDuvelle @WorldImagining @drkjjeffery @CogCompNeuro Hey something I wanted to ask. When I did a CCN keynote and tutorial, Kim and I were kind of ambushed with the tutorial request. Now that it is all in the open, do you get lots of good applicants? It's a huge commitment. I'm not sure I would have put myself forward if I'd known
2022-05-29 12:39:32 @usamakhawajaa Thanks!
2022-05-29 12:39:25 @usamakhawajaa Awesome - will take a look (after the grant is written :) )
2022-05-29 12:35:17 @usamakhawajaa As in fine tune the weights? Is that an option? I just have the API for the trained model.
2022-05-29 12:30:42 @Dick_retired @fmrib_steve Why do you use a computer to analyse your data? These models will be helping us think soon. So long as I am still a small part of the equation, I am good with that if the science gets more interesting :)
2022-05-29 09:22:17 @AthenaAkrami @AcerbiLuigi @TPVogels @biorxivpreprint I have been experimenting with a lab seal, but the wax keeps slipping down my screen.
2022-05-29 08:00:15 @GunnarBlohm @neuralreckoning @KordingLab @ElDuvelle @WorldImagining @drkjjeffery Working on it. More info later in the year :)
2022-05-29 07:50:11 @fmrib_steve Just totally shallow. I agree some grants are like that. Not the funded ones though. In my experience at least :)
2022-05-29 07:49:25 @fmrib_steve I was pretty confident about this one - I don't see any problem using GPT as a tool to help scientific writing though, if that's what your asking. Or even idea generation. Maybe some version can be useful in the future. Not this version though :)
2022-05-29 07:23:59 Nice try, but I don't think GPT-3 is going to get funded yet. https://t.co/rmzCxbgpkx
2022-05-28 19:21:36 @neuralreckoning @KordingLab @GunnarBlohm @ElDuvelle @WorldImagining @drkjjeffery Currently they are almost exclusively selecting mice :)
2022-05-28 19:20:02 @neuralreckoning @KordingLab @GunnarBlohm @ElDuvelle @WorldImagining @drkjjeffery They try to, but more often what they select becomes important (at least temporarily). Can be a good thing if done well. Or can cause Brexit if Murdoch decides immigration is important. They set the conversation, and we all argue about LTP for 3 decades.
2022-05-28 19:08:53 @neuralreckoning @KordingLab @GunnarBlohm @ElDuvelle @WorldImagining @drkjjeffery People read Nature etc. because that’s what the current conversation is about and they want to be involved. Same reason people read tatler. The whole industry is built on filtering of this sort. Whatever we make next needs to understand that. Scientists are humans too.
2022-05-28 09:28:42 This is such an exciting model. Awesome theory lab hires experimental postdocs ->
2022-05-27 15:04:40 RT @SaxeLab: Official advert for ~3 postdoctoral positions focused on theory-driven experiments now live:https://t.co/biN5ZKZS4a Undert…
2022-05-27 06:40:23 @neurograce @nyuniversity @NYUPsych @NYUDataScience Congrats!!
2022-05-27 06:36:38 RT @profgeraintrees: Not just the US. @UKRI_News doctoral stipends are below the living wage in the UK. And inflation is now a real challen…
2022-05-27 05:40:55 Here is the current board https://t.co/nVYFUIFy4S
2022-05-27 05:40:18 This is an amazing opportunity. The @elife board is made up of leadership figures from HHMI, Wellcome, Max Planck etc. but also includes scientists, including (usually 1?) ECR. Non-executive Director (Scientific) https://t.co/1cJl2MOY2X
2022-05-26 21:33:21 @tyrell_turing @ElDuvelle @neuralreckoning PS I think what Dan is doing is great and I’m looking forward to seeing the new neuromatch unjournal :)
2022-05-26 21:32:13 @tyrell_turing @ElDuvelle @neuralreckoning You’re assuming his contribution to elife was a net +ve
2022-05-26 17:57:53 RT @cOAlitionS_OA: We have some exciting news to share: the Journal Comparison Service (JCS) is now open for publishers to register and dep…
2022-05-26 03:52:15 RT @chingfang17: What biological mechanisms support learning of predictive maps? In our new preprint with D. Aronov, L. Abbott, &
2022-05-25 19:28:33 RT @rmnpogodin: We have a few positions at Gatsby for up to a year! Available for undergrads/master’s/PhD students
2022-05-25 13:47:48 This is @neuro_kim on steroids. I tried to keep up but totally failed. https://t.co/lQgCnSdBUx
2022-05-25 04:51:50 RT @AdamCSharp: Different types of animal poop and what they mean:7. Apeshit = angry6. Batshit = unhinged5. Bullshit = lies (or unfair)…
2022-05-25 04:22:13 RT @mameister4: Why does it take 10,000 trials to train a mouse on a trivial 2AFC task? When the same animal will learn a complex sequence…
2022-05-24 08:09:37 RT @LHuntNeuro: One week left to apply!
2022-05-22 15:22:41 RT @MargaretAtwood: How stupid is this? Or does uk leadership Want Uk to be a science &
2022-05-21 21:59:26 Think it might be time to limit screen time.6yo: “I don’t believe Jesus re-spawned, Daddy”
2022-05-20 08:11:00 CAFIAC FIX
2022-10-28 12:57:44 @drdevangm @PlantEvolution I'm glad @PlantEvolution returned home full of hope. @mbeisen and I returned home full of Gin. Was v nice to see you guys!
2022-10-28 08:33:29 @mbeisen I am still drunk.
2022-10-28 06:41:04 RT @TPVogels: We find ourselves in the curious position to have very few non-African applicants... Usually it's about ~100 applications fro…
2022-10-27 08:13:15 RT @MIcheleABasso1: Right on Marge!!Professor Livingstone’s personal statement about recent concerns over animal research | Harvard Medica…
2022-10-27 07:59:33 @NeilBurgess10 @drkjjeffery @neuralengine That doc I sent was the first one I think. Can dig out second one at some point. Have to go to elife meeting now though to see if we can resolve all these issues
2022-10-27 07:58:39 @NeilBurgess10 @drkjjeffery @neuralengine Trance = accept. Weird
2022-10-27 07:58:01 @NeilBurgess10 @drkjjeffery @neuralengine We are actually confusing two different trials. Sorry. First one triage plus guaranteed acceptance. Authors could withdraw. Second one no triage. There was a trance decision. But all reviews posted on biorxiv. No gender differences in people opting for either trial.
2022-10-27 07:29:59 @NeilBurgess10 @drkjjeffery @neuralengine Papers were not rejected after peer review. Authors knew we would post all reviews.
2022-10-27 07:28:17 @NeilBurgess10 @drkjjeffery @neuralengine No.
2022-10-27 07:21:26 @drkjjeffery @neuralengine @NeilBurgess10 https://t.co/EONCBOX9GT
2022-10-27 07:19:51 @drkjjeffery @neuralengine @NeilBurgess10 No this was a trial where all reviews were made public in exchange for guaranteed review (no triage). Everything we reviewed in the trial was accepted. Negative reviews published.
2022-10-27 05:33:59 @drkjjeffery @neuralengine @NeilBurgess10 We will be monitoring this but evidence to date suggests it isn’t true. Given a choice in our trial, women were no less likely to opt for public peer review than men. The only demographic less likely to prefer public peer review were late (compared to early) career researchers
2022-10-26 17:40:07 @AllenNeuroLab @stevenmweisberg @smfleming @eLife I **promise you** we’ll get to triage!!! There are more radical solutions than this on the table (as well as less radical ones)
2022-10-26 17:03:51 @AllenNeuroLab @stevenmweisberg @smfleming @eLife There is a real philosophical difference here. The new system is now like New York Times Film Reviews. You can't tell if the film is good just because the review is there. You need to read the assessment. But like NYT, we will still have to choice who we review, causing problems.
2022-10-26 17:01:08 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Some people are definitely shouting. Maybe not you :) We aren't moving peer review decisions to triage decisions. We are getting rid of the notion of acceptance being the goal.
2022-10-26 16:17:11 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Triage is on the agenda. We understand it's an issue. We will also try to figure out a way to do that better. In the meantime, I would be happy if people focused their criticisms on what we are doing with peer review.
2022-10-26 16:16:22 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We have been doing triage behind a barrier for a decade as do almost all journals. Now we change something totally different and everyone shouts at us about triage. It is just madness. It's like I told you I am moving house and you shouting at me because you don't like my work
2022-10-26 16:14:47 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We practically have to handle 7-900 papers a month in a relatively streamlined way. So any changes we do need to be gradual.
2022-10-26 16:13:00 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Not sure I really parse this message, but we can't change everything at once. eLife is like a massive oil tanker that takes a long time to turn. There are 650 editors who all have different views and practices and habits.
2022-10-26 15:23:52 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We know triage is a random and biased and hated thing. Can people let us try to solve one problem at a time please.
2022-10-23 19:58:08 @BorisLenhard @eLife https://t.co/a3SJJfUHqs
2022-10-23 19:39:39 @LilaDavachi @EAHellerPhD @PracheeAC @eLife Just upping this here as well as in DMs because I think people have missed it. Just noting that the more is a concrete evaluation as well as the peer reviews. We should def chat more about this though. https://t.co/a3SJJfDEos
2022-10-23 18:43:12 RT @JACoates: Can we just talk about the reality of peer review for a second because so many scientists here seem completely oblivous to th…
2022-10-23 18:39:38 @LilaDavachi @PracheeAC @eLife Hey Lila - I don't get what you are saying here, but we are really keen to understand the potential sources of bias in this new system that are different from the old. Is there any chance you could unpack it to explain more? Could be in DMs or email if easier
2022-10-23 18:37:33 RT @patrickmineault: Brains are cool! They are literally what make us, us! We should study them more, for their own sake! 6/6
2022-10-23 12:57:44 RT @LMT_Spoon: At the risk of wading into this debate very early in my career and appearing naive about the realities of scientific publish…
2022-10-23 12:57:34 RT @albertcardona: Wonderful to see an early career scientist with such a clear understanding of scientific publishing and what @eLife is t…
2022-10-23 10:19:39 @niallpboyce But how is that different to the current system? You guys are not transparent. You make mistakes. You publish stuff that doesn't replicate. You are entirely beholden to reviewer opinion. We are just saying this reviewer opinion is not binary. That's the only difference.
2022-10-23 09:50:25 @generegulation @eLife You haven't understood then. https://t.co/a3SJJfDEos
2022-10-23 09:45:41 @aazadmmn Haha - No I am just being childish. No subtext at all. Literally referring to the colour of a dead and living brain and the appearance of figures in ML papers. No depth at all. Just silliness. Trying to avoid any meaningful conversation about AI and neuro.
2022-10-23 09:41:16 @aazadmmn They don't seem to have a regular colour in my experience. Definitely not boxes though. When I see them they are usually more like rectangles with arrows pointing out of the ends. Sometimes trapeziums. Occasionally circles. Always arrows though.
2022-10-23 09:36:37 @aazadmmn Even the dead brain is at most a browny-grey box. Much easier. When it is up-and-running it is pinky-red. Piece of cake.
2022-10-23 09:34:04 @generegulation @eLife This is a really strange take. In either case the journal is endorsing the reviewers. Previously this was only indirect. Now direct. That is exactly what the "journal" brings. Endorsement of the assessment (not the paper).
2022-10-23 09:32:24 @generegulation @eLife The journal can put its reputation at stake by endorsing the reviews and assessment!!!
2022-10-22 18:02:14 Figuring out how the brain works seems hard enough to me. I'm happy to let someone else solve AI.
2022-10-22 15:23:01 @mazefire56 :) @jcrwhittington @dorrell_will
2022-10-22 15:12:48 @TonyZador I guess the first autoencoder that solved the variational inference problem was the Helmholtz machine (Hinton, Dayan, Senjowski) - a model of the visual system, trained by the wake-sleep algorithm. Although perhaps Zhaoping Li had an olfactory version earlier.
2022-10-22 14:53:19 @TonyZador Transformers are an extropolation of attention models ("Attention is all you need"), which were definitely inspired by human vision. Prob not neuro, but def psychology. eg https://t.co/WMpSNeoKNx
2022-10-22 11:12:00 @lam88324 @RafaCarazoSalas @micahgallen @vsbuffalo Ha - It's been 48 hours - give people a chance!! Almost every institution in the scientific world has signed a statement that they won't judge by IFs but they still do because there is no alternative. Let us try to build one, and join on this journey :)
2022-10-22 08:50:42 @PlantEvolution @fishspeciation @eLife @mbeisen I don't understand why people are assuming the worst from a charity run by people trying to improve things. In general with eLife ->
2022-10-22 08:47:18 @fishspeciation @PlantEvolution @eLife @mbeisen As far as I know we have never yet declined a fee waiver request. There is your data right there :)
2022-10-22 07:53:56 RT @DanGarrahan: The Brexit effect film is free to view. No remoaning - just cold, hard facts clearly explained by @FT writers and interv…
2022-10-22 07:09:18 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We totally agree with you that triage is a problem, and transparency is important. It’s on the agenda. One step at a time.
2022-10-22 07:08:25 @drkjjeffery @joshdubnau @WesterfieldJM @eLife Instead we are trying to change what it means to publish. To change the output of science evaluation to be the evaluation itself. We think this brings many benefits as laid out in the editorial. But it isn’t about transparency.
2022-10-22 07:06:28 @drkjjeffery @joshdubnau @WesterfieldJM @eLife This misunderstands what we are doing. One reason for public evaluation is transparency of decision. We agree this is important (and triage is a big part of the decision) but this is not what we are doing here.
2022-10-22 06:14:36 @BalsdonTarryn Correct stats are as follows ->
2022-10-22 06:12:24 @BalsdonTarryn Hi. This is inaccurate. Approx 50% of papers reviewed papers are rejected. Your stats are for papers that are invited for revision. We have always tried to ensure that after we invite a revision chances are great. Please correct. Thanks.
2022-10-22 06:09:00 RT @GatsbyUCL: Upcoming Faculty Openings!We intend to recruit one principal investigator in #TheoreticalNeuroscience and one in #Machine…
2022-10-22 06:04:58 @adredish How’s this for your CV.“This paper reports the landmark discovery that mouse hippocampal neurons rapidly play out different possible futures during a decision. This suggests a new role for hippocampus and a new mechanism for choice. The supporting evidence is compelling.”
2022-10-21 21:44:40 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell Oh I see. I think for the next 3 months triage will likely be similar as editors get used to the new system. A problem with a massive machine is that it takes a long time to turn.
2022-10-21 21:42:01 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell We will def publish more papers. We hope to *review* more papers too.
2022-10-21 21:41:21 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell What do you mean A vs B route?
2022-10-21 21:40:38 @ElDuvelle @AheadOfTheNerve It will be indexed in Pubmed so unfortunately it will get an impact factor. I wouldn’t be surprised if Thompson do not even recognise that there has been a change.
2022-10-21 21:38:46 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We also have a very high-touch system where many editors see each decision. I think this is important when reviews are public. Lots of layers of scrutiny. Hopefully this means we are less likely to screw it up.
2022-10-21 21:36:30 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We are a large scale trusted journal, and we are pivoting to public evaluation. That makes us unique I think.
2022-10-21 21:35:43 @drkjjeffery @joshdubnau @WesterfieldJM @eLife Nothing that others couldn’t do in time. But we have a massive head start - a huge machine of 650 editors who are field leaders committed to this. And we will do it all publicly with a serious attempt at consistency across the journal, and an eye to reviews that are useful.
2022-10-21 17:08:27 @AllenNeuroLab @stevenmweisberg @smfleming Yes it is not a peer reviewed journal. The papers are *already* published on biorxiv. We are not selecting which to publish. We are attaching a public assessment to some of them.
2022-10-21 17:05:42 @AllenNeuroLab @smfleming In the same way new york times decides which films to review. Yes there are biases here which we should worry about, but critically you can't tell whether the film is good without reading the review. It is very different when the decision itself does not signal quality.
2022-10-21 14:58:37 @RafaCarazoSalas @micahgallen @vsbuffalo @eLife The only way eLIfe dies is if it stops innovating. What else is it there for? Plenty of prestige journals.
2022-10-21 14:57:23 @ScottBrincat @vsbuffalo I think the reception has been mostly good. It's difficult to know what aspects to focus on, and different for different sectors. We decided to focus on the overall mission to start with then shift the focus to the details on the following days. We're moving slow :)
2022-10-21 14:11:29 RT @LnccBrown: as a long time senior editor for @eLife I (MJF) thought I'd weigh in. When I first heard about the plan in its initial conce…
2022-10-21 14:09:14 @joshdubnau @WesterfieldJM No. @elife badge will say nothing about the quality of the paper. It will promise a high quality review. That's the whole point. The paper is already pubilshed before we even see it. This whole thread misses this critical point.
2022-10-28 12:57:44 @drdevangm @PlantEvolution I'm glad @PlantEvolution returned home full of hope. @mbeisen and I returned home full of Gin. Was v nice to see you guys!
2022-10-28 08:33:29 @mbeisen I am still drunk.
2022-10-28 06:41:04 RT @TPVogels: We find ourselves in the curious position to have very few non-African applicants... Usually it's about ~100 applications fro…
2022-10-27 08:13:15 RT @MIcheleABasso1: Right on Marge!!Professor Livingstone’s personal statement about recent concerns over animal research | Harvard Medica…
2022-10-27 07:59:33 @NeilBurgess10 @drkjjeffery @neuralengine That doc I sent was the first one I think. Can dig out second one at some point. Have to go to elife meeting now though to see if we can resolve all these issues
2022-10-27 07:58:39 @NeilBurgess10 @drkjjeffery @neuralengine Trance = accept. Weird
2022-10-27 07:58:01 @NeilBurgess10 @drkjjeffery @neuralengine We are actually confusing two different trials. Sorry. First one triage plus guaranteed acceptance. Authors could withdraw. Second one no triage. There was a trance decision. But all reviews posted on biorxiv. No gender differences in people opting for either trial.
2022-10-27 07:29:59 @NeilBurgess10 @drkjjeffery @neuralengine Papers were not rejected after peer review. Authors knew we would post all reviews.
2022-10-27 07:28:17 @NeilBurgess10 @drkjjeffery @neuralengine No.
2022-10-27 07:21:26 @drkjjeffery @neuralengine @NeilBurgess10 https://t.co/EONCBOX9GT
2022-10-27 07:19:51 @drkjjeffery @neuralengine @NeilBurgess10 No this was a trial where all reviews were made public in exchange for guaranteed review (no triage). Everything we reviewed in the trial was accepted. Negative reviews published.
2022-10-27 05:33:59 @drkjjeffery @neuralengine @NeilBurgess10 We will be monitoring this but evidence to date suggests it isn’t true. Given a choice in our trial, women were no less likely to opt for public peer review than men. The only demographic less likely to prefer public peer review were late (compared to early) career researchers
2022-10-26 17:40:07 @AllenNeuroLab @stevenmweisberg @smfleming @eLife I **promise you** we’ll get to triage!!! There are more radical solutions than this on the table (as well as less radical ones)
2022-10-26 17:03:51 @AllenNeuroLab @stevenmweisberg @smfleming @eLife There is a real philosophical difference here. The new system is now like New York Times Film Reviews. You can't tell if the film is good just because the review is there. You need to read the assessment. But like NYT, we will still have to choice who we review, causing problems.
2022-10-26 17:01:08 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Some people are definitely shouting. Maybe not you :) We aren't moving peer review decisions to triage decisions. We are getting rid of the notion of acceptance being the goal.
2022-10-26 16:17:11 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Triage is on the agenda. We understand it's an issue. We will also try to figure out a way to do that better. In the meantime, I would be happy if people focused their criticisms on what we are doing with peer review.
2022-10-26 16:16:22 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We have been doing triage behind a barrier for a decade as do almost all journals. Now we change something totally different and everyone shouts at us about triage. It is just madness. It's like I told you I am moving house and you shouting at me because you don't like my work
2022-10-26 16:14:47 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We practically have to handle 7-900 papers a month in a relatively streamlined way. So any changes we do need to be gradual.
2022-10-26 16:13:00 @AllenNeuroLab @stevenmweisberg @smfleming @eLife Not sure I really parse this message, but we can't change everything at once. eLife is like a massive oil tanker that takes a long time to turn. There are 650 editors who all have different views and practices and habits.
2022-10-26 15:23:52 @AllenNeuroLab @stevenmweisberg @smfleming @eLife We know triage is a random and biased and hated thing. Can people let us try to solve one problem at a time please.
2022-10-23 19:58:08 @BorisLenhard @eLife https://t.co/a3SJJfUHqs
2022-10-23 19:39:39 @LilaDavachi @EAHellerPhD @PracheeAC @eLife Just upping this here as well as in DMs because I think people have missed it. Just noting that the more is a concrete evaluation as well as the peer reviews. We should def chat more about this though. https://t.co/a3SJJfDEos
2022-10-23 18:43:12 RT @JACoates: Can we just talk about the reality of peer review for a second because so many scientists here seem completely oblivous to th…
2022-10-23 18:39:38 @LilaDavachi @PracheeAC @eLife Hey Lila - I don't get what you are saying here, but we are really keen to understand the potential sources of bias in this new system that are different from the old. Is there any chance you could unpack it to explain more? Could be in DMs or email if easier
2022-10-23 18:37:33 RT @patrickmineault: Brains are cool! They are literally what make us, us! We should study them more, for their own sake! 6/6
2022-10-23 12:57:44 RT @LMT_Spoon: At the risk of wading into this debate very early in my career and appearing naive about the realities of scientific publish…
2022-10-23 12:57:34 RT @albertcardona: Wonderful to see an early career scientist with such a clear understanding of scientific publishing and what @eLife is t…
2022-10-23 10:19:39 @niallpboyce But how is that different to the current system? You guys are not transparent. You make mistakes. You publish stuff that doesn't replicate. You are entirely beholden to reviewer opinion. We are just saying this reviewer opinion is not binary. That's the only difference.
2022-10-23 09:50:25 @generegulation @eLife You haven't understood then. https://t.co/a3SJJfDEos
2022-10-23 09:45:41 @aazadmmn Haha - No I am just being childish. No subtext at all. Literally referring to the colour of a dead and living brain and the appearance of figures in ML papers. No depth at all. Just silliness. Trying to avoid any meaningful conversation about AI and neuro.
2022-10-23 09:41:16 @aazadmmn They don't seem to have a regular colour in my experience. Definitely not boxes though. When I see them they are usually more like rectangles with arrows pointing out of the ends. Sometimes trapeziums. Occasionally circles. Always arrows though.
2022-10-23 09:36:37 @aazadmmn Even the dead brain is at most a browny-grey box. Much easier. When it is up-and-running it is pinky-red. Piece of cake.
2022-10-23 09:34:04 @generegulation @eLife This is a really strange take. In either case the journal is endorsing the reviewers. Previously this was only indirect. Now direct. That is exactly what the "journal" brings. Endorsement of the assessment (not the paper).
2022-10-23 09:32:24 @generegulation @eLife The journal can put its reputation at stake by endorsing the reviews and assessment!!!
2022-10-22 18:02:14 Figuring out how the brain works seems hard enough to me. I'm happy to let someone else solve AI.
2022-10-22 15:23:01 @mazefire56 :) @jcrwhittington @dorrell_will
2022-10-22 15:12:48 @TonyZador I guess the first autoencoder that solved the variational inference problem was the Helmholtz machine (Hinton, Dayan, Senjowski) - a model of the visual system, trained by the wake-sleep algorithm. Although perhaps Zhaoping Li had an olfactory version earlier.
2022-10-22 14:53:19 @TonyZador Transformers are an extropolation of attention models ("Attention is all you need"), which were definitely inspired by human vision. Prob not neuro, but def psychology. eg https://t.co/WMpSNeoKNx
2022-10-22 11:12:00 @lam88324 @RafaCarazoSalas @micahgallen @vsbuffalo Ha - It's been 48 hours - give people a chance!! Almost every institution in the scientific world has signed a statement that they won't judge by IFs but they still do because there is no alternative. Let us try to build one, and join on this journey :)
2022-10-22 08:50:42 @PlantEvolution @fishspeciation @eLife @mbeisen I don't understand why people are assuming the worst from a charity run by people trying to improve things. In general with eLife ->
2022-10-22 08:47:18 @fishspeciation @PlantEvolution @eLife @mbeisen As far as I know we have never yet declined a fee waiver request. There is your data right there :)
2022-10-22 07:53:56 RT @DanGarrahan: The Brexit effect film is free to view. No remoaning - just cold, hard facts clearly explained by @FT writers and interv…
2022-10-22 07:09:18 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We totally agree with you that triage is a problem, and transparency is important. It’s on the agenda. One step at a time.
2022-10-22 07:08:25 @drkjjeffery @joshdubnau @WesterfieldJM @eLife Instead we are trying to change what it means to publish. To change the output of science evaluation to be the evaluation itself. We think this brings many benefits as laid out in the editorial. But it isn’t about transparency.
2022-10-22 07:06:28 @drkjjeffery @joshdubnau @WesterfieldJM @eLife This misunderstands what we are doing. One reason for public evaluation is transparency of decision. We agree this is important (and triage is a big part of the decision) but this is not what we are doing here.
2022-10-22 06:14:36 @BalsdonTarryn Correct stats are as follows ->
2022-10-22 06:12:24 @BalsdonTarryn Hi. This is inaccurate. Approx 50% of papers reviewed papers are rejected. Your stats are for papers that are invited for revision. We have always tried to ensure that after we invite a revision chances are great. Please correct. Thanks.
2022-10-22 06:09:00 RT @GatsbyUCL: Upcoming Faculty Openings!We intend to recruit one principal investigator in #TheoreticalNeuroscience and one in #Machine…
2022-10-22 06:04:58 @adredish How’s this for your CV.“This paper reports the landmark discovery that mouse hippocampal neurons rapidly play out different possible futures during a decision. This suggests a new role for hippocampus and a new mechanism for choice. The supporting evidence is compelling.”
2022-10-21 21:44:40 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell Oh I see. I think for the next 3 months triage will likely be similar as editors get used to the new system. A problem with a massive machine is that it takes a long time to turn.
2022-10-21 21:42:01 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell We will def publish more papers. We hope to *review* more papers too.
2022-10-21 21:41:21 @joshdubnau @WesterfieldJM @eLife @ScottishWaddell What do you mean A vs B route?
2022-10-21 21:40:38 @ElDuvelle @AheadOfTheNerve It will be indexed in Pubmed so unfortunately it will get an impact factor. I wouldn’t be surprised if Thompson do not even recognise that there has been a change.
2022-10-21 21:38:46 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We also have a very high-touch system where many editors see each decision. I think this is important when reviews are public. Lots of layers of scrutiny. Hopefully this means we are less likely to screw it up.
2022-10-21 21:36:30 @drkjjeffery @joshdubnau @WesterfieldJM @eLife We are a large scale trusted journal, and we are pivoting to public evaluation. That makes us unique I think.
2022-10-21 21:35:43 @drkjjeffery @joshdubnau @WesterfieldJM @eLife Nothing that others couldn’t do in time. But we have a massive head start - a huge machine of 650 editors who are field leaders committed to this. And we will do it all publicly with a serious attempt at consistency across the journal, and an eye to reviews that are useful.
2022-10-21 17:08:27 @AllenNeuroLab @stevenmweisberg @smfleming Yes it is not a peer reviewed journal. The papers are *already* published on biorxiv. We are not selecting which to publish. We are attaching a public assessment to some of them.
2022-10-21 17:05:42 @AllenNeuroLab @smfleming In the same way new york times decides which films to review. Yes there are biases here which we should worry about, but critically you can't tell whether the film is good without reading the review. It is very different when the decision itself does not signal quality.
2022-10-21 14:58:37 @RafaCarazoSalas @micahgallen @vsbuffalo @eLife The only way eLIfe dies is if it stops innovating. What else is it there for? Plenty of prestige journals.
2022-10-21 14:57:23 @ScottBrincat @vsbuffalo I think the reception has been mostly good. It's difficult to know what aspects to focus on, and different for different sectors. We decided to focus on the overall mission to start with then shift the focus to the details on the following days. We're moving slow :)
2022-10-21 14:11:29 RT @LnccBrown: as a long time senior editor for @eLife I (MJF) thought I'd weigh in. When I first heard about the plan in its initial conce…
2022-10-21 14:09:14 @joshdubnau @WesterfieldJM No. @elife badge will say nothing about the quality of the paper. It will promise a high quality review. That's the whole point. The paper is already pubilshed before we even see it. This whole thread misses this critical point.
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-21 19:18:58 @sainsburys this bottle still had the theft guard on when it came in our delivery. Order number 682672550 https://t.co/TIar9SIRY1
2022-11-21 10:06:00 RT @MedinaRiojaMD: Frontal Lobe Anatomy part 1: gross anatomy https://t.co/Gi2XPwcBNz
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-22 20:58:48 Really exciting new lab for high quality human work. Nils is a deep thinker. Great place to learn the whole gamut of human neuro skills. https://t.co/oHsFuVWkM4
2022-11-22 19:19:04 @JasonHise64 @dorrell_will @jcrwhittington I told you we should be studying rubix cubes experts!!
2022-11-22 19:15:49 RT @NilsKolling: Excited to announce my ERC StG @ERC_Research! This means I will be recruiting soon! As its a large grant, I will be lookin…
2022-11-22 18:25:43 @sainsburys Hi - I sent a DM on this link last night. Just checking you got it ok
2022-11-21 19:18:58 @sainsburys this bottle still had the theft guard on when it came in our delivery. Order number 682672550 https://t.co/TIar9SIRY1
2022-11-21 10:06:00 RT @MedinaRiojaMD: Frontal Lobe Anatomy part 1: gross anatomy https://t.co/Gi2XPwcBNz
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-24 07:10:38 @YardenJCohen The current system is just a ranking. If we agree we need a ranking, no reason a guide can’t produce one. There can be more than one guide too. Like with restaurants and films. Curation of science should not be tied in with publication.
2022-11-24 06:47:38 @YardenJCohen Reviewed preprints are the way forward. We basically need a preprint server plus a guide on top to tell us what to trust and what to read.
2022-11-23 17:24:39 Wouldn’t it be good if there was a journal where you could have an honest conversation with reviewers because they could comment but couldn’t stop you publishing? https://t.co/kjqjfMYMi0
2022-11-23 08:54:51 We'd like to thank the reviewer for this excellent suggestion. https://t.co/27tnUoEHmG
2022-11-23 08:42:52 @dorrell_will @JasonHise64 @jcrwhittington yeah come on @GabyMohamady @nikasamborska @BeatrizSimG @petertdoohan
2022-11-22 20:58:48 Really exciting new lab for high quality human work. Nils is a deep thinker. Great place to learn the whole gamut of human neuro skills. https://t.co/oHsFuVWkM4
2022-11-22 19:19:04 @JasonHise64 @dorrell_will @jcrwhittington I told you we should be studying rubix cubes experts!!
2022-11-22 19:15:49 RT @NilsKolling: Excited to announce my ERC StG @ERC_Research! This means I will be recruiting soon! As its a large grant, I will be lookin…
2022-11-22 18:25:43 @sainsburys Hi - I sent a DM on this link last night. Just checking you got it ok
2022-11-21 19:18:58 @sainsburys this bottle still had the theft guard on when it came in our delivery. Order number 682672550 https://t.co/TIar9SIRY1
2022-11-21 10:06:00 RT @MedinaRiojaMD: Frontal Lobe Anatomy part 1: gross anatomy https://t.co/Gi2XPwcBNz
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-25 19:58:22 @ElDuvelle @drkjjeffery @PierreY83952112 We’re kind of hoping they *won’t* be there in Hc but *will* be there in CTX
2022-11-25 19:44:33 @_AndyAlexander @AdrianDu_ @ElDuvelle @PierreY83952112 @nitz_douglas This is exactly the ticket!
2022-11-25 19:44:01 @_AndyAlexander @AdrianDu_ @ElDuvelle @PierreY83952112 @nitz_douglas Ooh - awesome!!!
2022-11-25 19:26:05 @samamckenzie @drkjjeffery @ElDuvelle @PierreY83952112 Brilliant. Thanks Sam!
2022-11-25 19:25:19 @AdrianDu_ @ElDuvelle @PierreY83952112 @_AndyAlexander @nitz_douglas Yes this looks really really relevant. Thank you!!!
2022-11-25 19:06:08 @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas @colinlever Yes these are a perfect example. Definitely abstract for our purpose because they generalise over objects. We know of these ones (and bvcs)
2022-11-25 19:04:51 @drkjjeffery @ElDuvelle @PierreY83952112 Matched in and out of corridors for speed etc..
2022-11-25 19:04:31 @drkjjeffery @ElDuvelle @PierreY83952112 Bit hard to go into details of the model here, but you can definitely separate from basic actions in data,no?. Eg door cell above, the actions and head direction cells will all be different for different doors. Eg corridors can be oriented differently and running can be…
2022-11-25 18:57:53 @TheBrunoCortex @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas Super cool. Thanks Randy. Would be great to meet you now you’re in Ox. Will be in DPAG for a lecture on 5th Dec if you’re around?
2022-11-25 18:49:01 @drkjjeffery @ElDuvelle @PierreY83952112 And yes the policies that become attached to these abstract features do look sometimes like your symmetric direction cells :)
2022-11-25 18:48:00 @drkjjeffery @ElDuvelle @PierreY83952112 Anything where a particular spatial layout always implies a common policy (go through doors. Run along corridors. Run around corners. Obvs run along boundaries is a given). In mazes these could be quite counter-intuitive (eg run around “u” of “w-maze”)..
2022-11-25 18:29:59 RT @ElDuvelle: Anyone can help Tim with examples of abstraction-like neural encoding? Probably in retrosplenial or mPCF? But elsewhere is…
2022-11-25 18:22:27 @ElDuvelle @PierreY83952112 Ps thanks!!!
2022-11-25 18:19:46 @ElDuvelle @PierreY83952112 Yep. I was thinking retrosplenial/mpfc. And yep more like objects or common spatial patterns etc.
2022-11-25 18:04:09 @ElDuvelle @PierreY83952112 Hi El - font of all knowledge - Do you know of other cool cells like this but for different abstractions (like corners or corridors etc.)?? We have a preliminary abstraction model but no data to predict.
2022-11-25 05:57:01 RT @JoshMitte: 1. Universities will be banned from greater income to provide better services 2. UK will be deprived from top minds by preve…
2022-11-25 05:53:19 RT @KeithAdamTaylor: @AllieHBNews As an attendee of many trade fairs across Asia, the two things the UK advertises at these events are univ…
2022-11-24 07:10:38 @YardenJCohen The current system is just a ranking. If we agree we need a ranking, no reason a guide can’t produce one. There can be more than one guide too. Like with restaurants and films. Curation of science should not be tied in with publication.
2022-11-24 06:47:38 @YardenJCohen Reviewed preprints are the way forward. We basically need a preprint server plus a guide on top to tell us what to trust and what to read.
2022-11-23 17:24:39 Wouldn’t it be good if there was a journal where you could have an honest conversation with reviewers because they could comment but couldn’t stop you publishing? https://t.co/kjqjfMYMi0
2022-11-23 08:54:51 We'd like to thank the reviewer for this excellent suggestion. https://t.co/27tnUoEHmG
2022-11-23 08:42:52 @dorrell_will @JasonHise64 @jcrwhittington yeah come on @GabyMohamady @nikasamborska @BeatrizSimG @petertdoohan
2022-11-22 20:58:48 Really exciting new lab for high quality human work. Nils is a deep thinker. Great place to learn the whole gamut of human neuro skills. https://t.co/oHsFuVWkM4
2022-11-22 19:19:04 @JasonHise64 @dorrell_will @jcrwhittington I told you we should be studying rubix cubes experts!!
2022-11-22 19:15:49 RT @NilsKolling: Excited to announce my ERC StG @ERC_Research! This means I will be recruiting soon! As its a large grant, I will be lookin…
2022-11-22 18:25:43 @sainsburys Hi - I sent a DM on this link last night. Just checking you got it ok
2022-11-21 19:18:58 @sainsburys this bottle still had the theft guard on when it came in our delivery. Order number 682672550 https://t.co/TIar9SIRY1
2022-11-21 10:06:00 RT @MedinaRiojaMD: Frontal Lobe Anatomy part 1: gross anatomy https://t.co/Gi2XPwcBNz
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-29 00:39:28 RT @SaxeLab: Excited to share our new paper at #NeurIPS2022! Prior knowledge deeply impacts learning dynamics. We derive exact solutions…
2022-11-28 06:43:54 @fmrib_steve CRISPR is amazing!
2022-11-27 17:58:00 RT @EikoFried: I dedicated my PhD to Aaron’s memory. If you have not heard about his work or life please take a moment to take a look below…
2022-11-26 15:52:18 @DLBarack @ElDuvelle @PierreY83952112 @drkjjeffery Haha! Philosophers have argued about everything!!
2022-11-26 15:45:16 @ElDuvelle @drkjjeffery @PierreY83952112 Thanks Kate!!
2022-11-26 15:09:06 @ElDuvelle @PierreY83952112 @drkjjeffery yes!!
2022-11-26 14:03:52 @ElDuvelle @PierreY83952112 @drkjjeffery Related q->
2022-11-26 08:18:25 @nicognitive @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas @colinlever Thank you :)
2022-11-25 19:58:22 @ElDuvelle @drkjjeffery @PierreY83952112 We’re kind of hoping they *won’t* be there in Hc but *will* be there in CTX
2022-11-25 19:44:33 @_AndyAlexander @AdrianDu_ @ElDuvelle @PierreY83952112 @nitz_douglas This is exactly the ticket!
2022-11-25 19:44:01 @_AndyAlexander @AdrianDu_ @ElDuvelle @PierreY83952112 @nitz_douglas Ooh - awesome!!!
2022-11-25 19:26:05 @samamckenzie @drkjjeffery @ElDuvelle @PierreY83952112 Brilliant. Thanks Sam!
2022-11-25 19:25:19 @AdrianDu_ @ElDuvelle @PierreY83952112 @_AndyAlexander @nitz_douglas Yes this looks really really relevant. Thank you!!!
2022-11-25 19:06:08 @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas @colinlever Yes these are a perfect example. Definitely abstract for our purpose because they generalise over objects. We know of these ones (and bvcs)
2022-11-25 19:04:51 @drkjjeffery @ElDuvelle @PierreY83952112 Matched in and out of corridors for speed etc..
2022-11-25 19:04:31 @drkjjeffery @ElDuvelle @PierreY83952112 Bit hard to go into details of the model here, but you can definitely separate from basic actions in data,no?. Eg door cell above, the actions and head direction cells will all be different for different doors. Eg corridors can be oriented differently and running can be…
2022-11-25 18:57:53 @TheBrunoCortex @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas Super cool. Thanks Randy. Would be great to meet you now you’re in Ox. Will be in DPAG for a lecture on 5th Dec if you’re around?
2022-11-25 18:49:01 @drkjjeffery @ElDuvelle @PierreY83952112 And yes the policies that become attached to these abstract features do look sometimes like your symmetric direction cells :)
2022-11-25 18:48:00 @drkjjeffery @ElDuvelle @PierreY83952112 Anything where a particular spatial layout always implies a common policy (go through doors. Run along corridors. Run around corners. Obvs run along boundaries is a given). In mazes these could be quite counter-intuitive (eg run around “u” of “w-maze”)..
2022-11-25 18:29:59 RT @ElDuvelle: Anyone can help Tim with examples of abstraction-like neural encoding? Probably in retrosplenial or mPCF? But elsewhere is…
2022-11-25 18:22:27 @ElDuvelle @PierreY83952112 Ps thanks!!!
2022-11-25 18:19:46 @ElDuvelle @PierreY83952112 Yep. I was thinking retrosplenial/mpfc. And yep more like objects or common spatial patterns etc.
2022-11-25 18:04:09 @ElDuvelle @PierreY83952112 Hi El - font of all knowledge - Do you know of other cool cells like this but for different abstractions (like corners or corridors etc.)?? We have a preliminary abstraction model but no data to predict.
2022-11-25 05:57:01 RT @JoshMitte: 1. Universities will be banned from greater income to provide better services 2. UK will be deprived from top minds by preve…
2022-11-25 05:53:19 RT @KeithAdamTaylor: @AllieHBNews As an attendee of many trade fairs across Asia, the two things the UK advertises at these events are univ…
2022-11-24 07:10:38 @YardenJCohen The current system is just a ranking. If we agree we need a ranking, no reason a guide can’t produce one. There can be more than one guide too. Like with restaurants and films. Curation of science should not be tied in with publication.
2022-11-24 06:47:38 @YardenJCohen Reviewed preprints are the way forward. We basically need a preprint server plus a guide on top to tell us what to trust and what to read.
2022-11-23 17:24:39 Wouldn’t it be good if there was a journal where you could have an honest conversation with reviewers because they could comment but couldn’t stop you publishing? https://t.co/kjqjfMYMi0
2022-11-23 08:54:51 We'd like to thank the reviewer for this excellent suggestion. https://t.co/27tnUoEHmG
2022-11-23 08:42:52 @dorrell_will @JasonHise64 @jcrwhittington yeah come on @GabyMohamady @nikasamborska @BeatrizSimG @petertdoohan
2022-11-22 20:58:48 Really exciting new lab for high quality human work. Nils is a deep thinker. Great place to learn the whole gamut of human neuro skills. https://t.co/oHsFuVWkM4
2022-11-22 19:19:04 @JasonHise64 @dorrell_will @jcrwhittington I told you we should be studying rubix cubes experts!!
2022-11-22 19:15:49 RT @NilsKolling: Excited to announce my ERC StG @ERC_Research! This means I will be recruiting soon! As its a large grant, I will be lookin…
2022-11-22 18:25:43 @sainsburys Hi - I sent a DM on this link last night. Just checking you got it ok
2022-11-21 19:18:58 @sainsburys this bottle still had the theft guard on when it came in our delivery. Order number 682672550 https://t.co/TIar9SIRY1
2022-11-21 10:06:00 RT @MedinaRiojaMD: Frontal Lobe Anatomy part 1: gross anatomy https://t.co/Gi2XPwcBNz
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife
2022-11-29 00:39:28 RT @SaxeLab: Excited to share our new paper at #NeurIPS2022! Prior knowledge deeply impacts learning dynamics. We derive exact solutions…
2022-11-28 06:43:54 @fmrib_steve CRISPR is amazing!
2022-11-27 17:58:00 RT @EikoFried: I dedicated my PhD to Aaron’s memory. If you have not heard about his work or life please take a moment to take a look below…
2022-11-26 15:52:18 @DLBarack @ElDuvelle @PierreY83952112 @drkjjeffery Haha! Philosophers have argued about everything!!
2022-11-26 15:45:16 @ElDuvelle @drkjjeffery @PierreY83952112 Thanks Kate!!
2022-11-26 15:09:06 @ElDuvelle @PierreY83952112 @drkjjeffery yes!!
2022-11-26 14:03:52 @ElDuvelle @PierreY83952112 @drkjjeffery Related q->
2022-11-26 08:18:25 @nicognitive @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas @colinlever Thank you :)
2022-11-25 19:58:22 @ElDuvelle @drkjjeffery @PierreY83952112 We’re kind of hoping they *won’t* be there in Hc but *will* be there in CTX
2022-11-25 19:44:33 @_AndyAlexander @AdrianDu_ @ElDuvelle @PierreY83952112 @nitz_douglas This is exactly the ticket!
2022-11-25 19:44:01 @_AndyAlexander @AdrianDu_ @ElDuvelle @PierreY83952112 @nitz_douglas Ooh - awesome!!!
2022-11-25 19:26:05 @samamckenzie @drkjjeffery @ElDuvelle @PierreY83952112 Brilliant. Thanks Sam!
2022-11-25 19:25:19 @AdrianDu_ @ElDuvelle @PierreY83952112 @_AndyAlexander @nitz_douglas Yes this looks really really relevant. Thank you!!!
2022-11-25 19:06:08 @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas @colinlever Yes these are a perfect example. Definitely abstract for our purpose because they generalise over objects. We know of these ones (and bvcs)
2022-11-25 19:04:51 @drkjjeffery @ElDuvelle @PierreY83952112 Matched in and out of corridors for speed etc..
2022-11-25 19:04:31 @drkjjeffery @ElDuvelle @PierreY83952112 Bit hard to go into details of the model here, but you can definitely separate from basic actions in data,no?. Eg door cell above, the actions and head direction cells will all be different for different doors. Eg corridors can be oriented differently and running can be…
2022-11-25 18:57:53 @TheBrunoCortex @ElDuvelle @PierreY83952112 @drkjjeffery @drzhangningyu @nitz_douglas Super cool. Thanks Randy. Would be great to meet you now you’re in Ox. Will be in DPAG for a lecture on 5th Dec if you’re around?
2022-11-25 18:49:01 @drkjjeffery @ElDuvelle @PierreY83952112 And yes the policies that become attached to these abstract features do look sometimes like your symmetric direction cells :)
2022-11-25 18:48:00 @drkjjeffery @ElDuvelle @PierreY83952112 Anything where a particular spatial layout always implies a common policy (go through doors. Run along corridors. Run around corners. Obvs run along boundaries is a given). In mazes these could be quite counter-intuitive (eg run around “u” of “w-maze”)..
2022-11-25 18:29:59 RT @ElDuvelle: Anyone can help Tim with examples of abstraction-like neural encoding? Probably in retrosplenial or mPCF? But elsewhere is…
2022-11-25 18:22:27 @ElDuvelle @PierreY83952112 Ps thanks!!!
2022-11-25 18:19:46 @ElDuvelle @PierreY83952112 Yep. I was thinking retrosplenial/mpfc. And yep more like objects or common spatial patterns etc.
2022-11-25 18:04:09 @ElDuvelle @PierreY83952112 Hi El - font of all knowledge - Do you know of other cool cells like this but for different abstractions (like corners or corridors etc.)?? We have a preliminary abstraction model but no data to predict.
2022-11-25 05:57:01 RT @JoshMitte: 1. Universities will be banned from greater income to provide better services 2. UK will be deprived from top minds by preve…
2022-11-25 05:53:19 RT @KeithAdamTaylor: @AllieHBNews As an attendee of many trade fairs across Asia, the two things the UK advertises at these events are univ…
2022-11-24 07:10:38 @YardenJCohen The current system is just a ranking. If we agree we need a ranking, no reason a guide can’t produce one. There can be more than one guide too. Like with restaurants and films. Curation of science should not be tied in with publication.
2022-11-24 06:47:38 @YardenJCohen Reviewed preprints are the way forward. We basically need a preprint server plus a guide on top to tell us what to trust and what to read.
2022-11-23 17:24:39 Wouldn’t it be good if there was a journal where you could have an honest conversation with reviewers because they could comment but couldn’t stop you publishing? https://t.co/kjqjfMYMi0
2022-11-23 08:54:51 We'd like to thank the reviewer for this excellent suggestion. https://t.co/27tnUoEHmG
2022-11-23 08:42:52 @dorrell_will @JasonHise64 @jcrwhittington yeah come on @GabyMohamady @nikasamborska @BeatrizSimG @petertdoohan
2022-11-22 20:58:48 Really exciting new lab for high quality human work. Nils is a deep thinker. Great place to learn the whole gamut of human neuro skills. https://t.co/oHsFuVWkM4
2022-11-22 19:19:04 @JasonHise64 @dorrell_will @jcrwhittington I told you we should be studying rubix cubes experts!!
2022-11-22 19:15:49 RT @NilsKolling: Excited to announce my ERC StG @ERC_Research! This means I will be recruiting soon! As its a large grant, I will be lookin…
2022-11-22 18:25:43 @sainsburys Hi - I sent a DM on this link last night. Just checking you got it ok
2022-11-21 19:18:58 @sainsburys this bottle still had the theft guard on when it came in our delivery. Order number 682672550 https://t.co/TIar9SIRY1
2022-11-21 10:06:00 RT @MedinaRiojaMD: Frontal Lobe Anatomy part 1: gross anatomy https://t.co/Gi2XPwcBNz
2022-11-19 07:08:55 Ooh exciting! Come on neuro Twitter we can do this! NIH vision neuroscientist. Male. Went to SfN. Writing paper on consciousness… https://t.co/dsxxB461SA
2022-11-18 20:31:19 RT @Li_Zhaoping: @behrenstimb A bit of self-advertising, my V1SH theory has more predictions (and exp. confirmed) than Barlow/Rao/Ballard/O…
2022-11-18 20:31:17 RT @Li_Zhaoping: @behrenstimb perhaps time to start a "vision" list :), for the new twitter. from your list: Bayes, Cajal, Helmholtz, Hubel…
2022-11-18 20:07:39 @Li_Zhaoping Rumours that 80% of the staff have left. Some who have left are saying that the it may go down irretrievably :).
2022-11-18 16:22:58 One last chance to see this? https://t.co/zHe4FaIEPJ
2022-11-18 12:50:40 RT @CKerren: The paper explaining the study, in which we used MEG to test the 2006 model by @Neurofool and @ptoncompmemlab on how the huma…
2022-11-17 10:48:06 @Phillips_M_G When I tried something similar, it got all the stereotyped stuff ok (loading the libraries), but the moment it had to do something remotely intelligent it failed dismally. eg it tried to make a maze using the connectivity matrix below. Fail. https://t.co/WAaRTqM8k7
2022-11-17 10:01:55 @caswellcaswell It’s got that right at least!!!
2022-11-17 10:01:43 RT @caswellcaswell: @behrenstimb Close but no cigar https://t.co/pLPRKTWCMO
2022-11-17 08:50:57 @SimeenTabassi @fmrib_steve Fake news!!! In all seriousness, although this is the answer I would give down the pub, it is disappointing that Google does not credit the whole team.
2022-11-17 08:19:26 This experience, more than any other with LLMs, convinces me that we need to separate the knowledge graph from the computations that use it. https://t.co/VQu1qQuhPi
2022-11-16 13:45:10 @gershbrain @nathanieldaw It almost nailed this one though.. https://t.co/pCgT1Ihfk7
2022-11-16 13:43:34 Nobel Prize here I come. https://t.co/6UqVGcxxq7
2022-11-16 09:40:55 @nathanieldaw :) This was my first foray into the field, so I kind of think I can be excused. I can't imagine what the dumb-ass reviewers were thinking though.
2022-11-16 09:38:57 @nathanieldaw Indeed :) Galactica's abstract is definitely more appropriate for the title. It would have been fine to just write that abstract and ignore the paper. I guess the worrying thing is that it cites the paper for the wrong thing.
2022-11-16 09:08:25 Perhaps most concerning, if I get it to write one of my papers, it realises the paper exists, and cites it, but gets the summary totally wrong. Instead it makes a plausible thing that it might be saying given the title. Not what the paper is actually saying at all. https://t.co/pLinV17EtY
2022-11-16 09:08:23 It can write a plausible abstract about why you want to model frontal cortex https://t.co/D1tYSjoOsp
2022-11-16 09:08:21 Not sure what happened here: https://t.co/EQbGZMdv9Q
2022-11-16 09:08:20 It does a better job on this one. https://t.co/2wF6ydi7tG
2022-11-16 09:08:18 Super fun. Early impressions are you need to be careful. Don't think we need to look for new jobs yet. https://t.co/lHR8o5NNID https://t.co/GJziSkJwjP
2022-11-15 20:12:54 RT @mbeisen: https://t.co/SWDb1VO4Wb @HHMINEWS @eLife