dezhejin.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

Out now in Nature Comms. To learn a new word, we need to remember it. We track factors driving memory of novel words, showing which words we remember or forget is predictable across people, and isolate a distinct region of fusiform cortex sensitive to this memorability. 🧠📈 #VisionScience 🧠💬 🧵👇

submitted 2 days ago • 2 comments

JAX with GPU is lightening faster than on CPU. Not even comparable.

submitted 1 day ago • 0 comments

Proud to work in this organization that puts unwavering support of its communities, open, independent science and scientific integrity front and center

submitted 5 days ago • 0 comments

Want to hear more about how feedback can guide learning in RNNs for motor adaptation. Here is our new paper in Nat. Com. with Barbara Feulner and @juangallego.bsky.social www.nature.com/articles/s41...

submitted 5 days ago • 0 comments

Friend of the pod Ross Taylor (ex reasoning lead for Llama) launched a platform for open reasoning data, gr.inc, seems promising!

submitted 5 days ago • 4 comments

Re-posting is appreciated: We have a fully funded PhD position in CMC lab @cmc-lab.bsky.social (at @tudresden_de). You can use forms.gle/qiAv5NZ871kv... to send your application and find more information. Deadline is April 30. Find more about CMC lab: cmclab.org and email me if you have questions.

submitted 6 days ago • 2 comments

Brains can be more than just a source of inspiration for AI capabilities; they can be a source of inspiration for AI safety, writes @patrickmineault.bsky.social in the latest entry in our NeuroAI series. www.thetransmitter.org/neuroai/does... #neuroskyence #NeuroAI #neuroscience

submitted 8 days ago • 2 comments

𝗕𝗿𝗮𝗶𝗻 𝗱𝘆𝗻𝗮𝗺𝗶𝗰𝘀? If interested in brain dynamics this one is a feast! www.nature.com/articles/s41... #neuroscience

submitted 8 days ago • 1 comment

If you've opened a new neuroscience lab recently, reach out! We're highlighting new PIs on the Liftoff series and our Launch: early career newsletter. You can sign up for the newsletter here: bit.ly/3X1SojT #neuroskyence

submitted 13 days ago • 0 comments

science.psu.edu/news/Jin2-2025 News on our recent paper published in Journal of Neuroscience. Birdsong syntax can be accurately inferred using statistical tests. The model can be used as a “language model” for Bengalese finch songs.

submitted 13 days ago • 0 comments

one benefit of open models is privacy 🔐you can run them locally & keep all your data on your local device instead of sending it to a company elsewhere congrats @soldaini.net for heavy lifting, showing our OLMoE model can run on iPhones📱

submitted 15 days ago • 0 comments

I remember saying in 2018 that it was a shame for those real scientists elected that year (including Peter Dayan) that it happened the same year the Royal Society killed its reputation by electing Elon Musk. The Royal Society can and should undo this disastrous mistake. Sign Stephen's letter below.

submitted 15 days ago • 0 comments

Our new work investigating multiplexing and mixed selectivity in auditory inferior colliculus neurons is up on bioRxiv 🎉 doi.org/10.1101/2025...

submitted 15 days ago • 0 comments

I'm giving an online talk tomorrow as part of UCL's NeuroAI series on neural architectures and our current line of research trying to figure out what they might be good for (including some philosophy: what might an answer to this question even look like?). www.eventbrite.co.uk/e/ucl-neuroa...

submitted 15 days ago • 0 comments

1/19 I’m thrilled that my postdoctoral work, with John Widloski and David Foster is now out in @science.org, along with a wonderful preview by Daniel Bendor!

submitted 25 days ago • 9 comments

Open AI dropped a new model, O3-mini-High. It is far superior than O1, Google Gemini 2.0 experimental, and DeepSeek R1, in that order. Open AI has some secret sauce. Can’t wait to see what O3 model will do!

submitted 25 days ago • 0 comments

This is rather strange. The Deepseek model was distilled and further trained for specific tasks, like O1 mini. As such, it is good at certain tasks, especially the benchmarks. When challenged with real-world problems, it often says oops, and excuses itself. It is specialized model, inferior to O1

submitted 28 days ago • 0 comments

Academic workers across the country are organizing to call on our congressional representatives on Thursday, 1/30, at 3pm ET / 12pm PT to demand these restrictions be lifted immediately. There will be a training over zoom at the beginning of the event. Join us! form.jotform.com/250226137228...

submitted 29 days ago • 17 comments

Back to work, back to #fluorescencefriday. Cheeky neurons in a mice hipoccampus #corefacility #lightmicroscopy

submitted 40 days ago • 0 comments

EXTREMELY cool and demonstrates clearly the promise of deep learning + bio. www.nature.com/articles/s41...

submitted 41 days ago • 4 comments

Clever study that explores the debate between those who claim that color categories are 'innate' and those who argue that they are dependent on language. It turns out that monkeys, unlike humans, do not have consensus color categories, suggesting cognitive mechanisms such as language are required.🧪🧠

submitted 41 days ago • 6 comments

Cerebellum is often systematically excluded from studies on brain, but it has >50% of neurons (yet belittled as “little brain”) & helps carry out work of body via spinal cord links, plus feedback loops w/ forebrain. Put at back, majority, doing extra work, underestimated - neural parallel to sexism.

submitted 42 days ago • 4 comments

Giving a talk tomorrow on "open science" practices including data sharing. Looking forward to sharing this lovely quote from @talyarkoni.com from a discussion on why we might share data and code (Tal, let me know if you object)

submitted 42 days ago • 2 comments

Way to go! Hope more will follow.

submitted 46 days ago • 0 comments

It’s an incredible feeling to see something you’ve poured your heart into come to life. This work was born from my curiosity about hippocampal “cognitive tuning.” Our interpretation of the environment depends on our actions, constantly evolving as we pursue our goals www.nature.com/articles/s41...

submitted 47 days ago • 6 comments

Synaptic chain network in mouse layer 2/3 motor cortex? Stimulating early active neurons pattern completes neural activity to drive movement. Stimulating late active neurons does not. Quite similar to what songbird HVC does! www.biorxiv.org/content/10.1...

submitted 47 days ago • 0 comments

Nature Hippocampal neuronal activity is aligned with action plans www.nature.com/articles/s41...

submitted 48 days ago • 0 comments

We value things differently depending on the situation. A fascinating paper shows that the hippocampus feeds contextual information to the prefrontal cortex by altering theta wave phase alignment, enabling the PFC to adjust value judgments accordingly. www.nature.com/articles/s41... #neuroscience

submitted 47 days ago • 5 comments

X who? Hope this is the year that it becomes where only zombies live.

submitted 50 days ago • 0 comments

x X has ~18m daily USA users. @bsky.app, ~1.5m. Is it worth 20 MINS a day to proactively recruit and bring your friends, neighbors,relatives, businesses, anyone who is or could be a user, to @bsky.app? If Bluesky ends up with more users than the others, you destabilize musk, zuck, trump.

submitted 50 days ago • 2894 comments

Some advice, as an eLife editor. (1) Don't send you Nature papers to eLife. (2) Do send your serious, full length studies. (3) You won't get a journal level IF to brag about but you will get the best peer review in the business and it will be seen by readers and will add value to your work.

submitted 76 days ago • 1 comment

Fantastic piece on the advances of NIH's BRAIN Initiative and the importance of protecting funding for basic circuits and systems neuroscience research! rdcu.be/d3g5R

submitted 76 days ago • 0 comments

From our recent paper: 7-color Tetbow image of the neocortex🌈 We can distinguish hundreds of neurons based on the combinatorial expression of 7 FPs. Here is the list of plasmids for Tetbow experiments if you are interested (all available from Addgene): sites.google.com/site/seedbre...

submitted 77 days ago • 3 comments

Nature Neuroscience is hiring a computational/systems/cognitive editor! Come join our team! Happy to answer any questions. springernature.wd3.myworkdayjobs.com/SpringerNatu...

submitted 79 days ago • 2 comments

Transformers operate in a block universe where all previous time steps are "simultaneously" available (time is spatialized). Like the brain, RNNs operate in a presentist universe in which only the current input is available and computations unfold in continuous time. For the record I'm a presentist.

submitted 79 days ago • 1 comment

Help bluesky neurohivemind! If you know of any computational / theoretical work modelling neuromodulators please share it 🙏 if you don't, please retweet!

submitted 80 days ago • 13 comments

"If Bluesky turns out to be a long-term home for political conversation, it will be because, as also happened on early Twitter, the new arrivals create appealing new habits and norms, not just new content," writes the author Clay Shirky. nyti.ms/4f2A28E

submitted 80 days ago • 28 comments

Fern leaf emerging from simple rules. This is a fun example I show at the end of the semester of Classical Mechanics. Behind complex phenomena, there might be simple reasons.

submitted 82 days ago • 0 comments

I'm searching for a research assistant to help with our new "models of mood" effort. Great fit for someone pre-PhD. If you know someone who might be interested, please pass this job ad along - thanks! wd1.myworkdaysite.com/en-US/recrui... What effort? This: www.nicolerust.com/thepivot

submitted 83 days ago • 4 comments

We are now up to 78 accounts on our Black In Neuro starter pack! Be sure to check back and make sure you have everyone added 🤗 Please let us know if we missed you! #BlackInSTEM #Neuroskyence #Psychology #AcademicSky 🧪 go.bsky.app/F8PpdPe

submitted 84 days ago • 11 comments

Are you interested in pursuing a PhD at the interface between Ai and Systems Neuroscience. If so: www.ai4biomed.io/research/pro... - a collaborative venture between UKRI Ai CDT in Biomedical Innovation and Simons Initiative for the Developing Brain (www.sidb.org.uk ) at The University of Edinburgh.

submitted 84 days ago • 0 comments

the rapid transition of academics off x (despite temporarily reducing reach/followers) makes you wonder what’s stopping us from ending the for-profit, closed access publishing industry. it’s, like…. we can just do it? or if not, interesting to consider what the inertial differences are.

submitted 88 days ago • 14 comments

How do LLMs learn to reason from data? Are they ~retrieving the answers from parametric knowledge🦜? In our new preprint, we look at the pretraining data and find evidence against this: Procedural knowledge in pretraining drives LLM reasoning ⚙️🔢 🧵⬇️

submitted 98 days ago • 38 comments

Just put together a starter pack for Deep Learning Theory. Let me know if you'd like to be included or suggest someone to add to the list! go.bsky.app/2qnppia

submitted 95 days ago • 29 comments