Weinan Sun

Thanks Misha for the shout-out!
It has been a truly mind-opening journey. WHOLISTIC is one step towards whole-organism neuroscience and a true physiological observatory - with extensions on the way! I am continually amazed by what the data reveals - so much remains to be discovered and mined.

Emergence of orthogonal hippocampal representations during spatial learning Learning & Behavior - Sun et al. (2025) reveal the progressive, dynamic changes in the response properties of thousands of hippocampal neurons as mice learn a conditional discrimination while...

New #Outlook #PsynomLB 'Emergence of orthogonal hippocampal representations during spatial learning' https://bit.ly/4mlOXQa by Prof. Vern Bingman @BGSU_Biology about the recent paper https://bit.ly/4jZO0eT by @sunw37 @JohanWinn @CornellNBB et al. @nature @Psychonomic_Soc

Excited to share our latest story! We found disentangled memory representations in the hippocampus that generalized across time and environments, despite the seemingly random drift and remapping of single cells. This code enabled the transfer of prior knowledge to solve new tasks

It was fun to visit Weinan's lab @sunw37.bsky.social at Cornell to give this talk. Exciting to see the experiments he's cooking up, and we had lots of fun discussions about PFC, hippocampus, consciousness and meditation!

Part of the talk content is here: blog.dileeplearning.com/p/space-is-a...

Large Concept Models: Language Modeling in a Sentence Representation Space – Janelia CVML Summary of and thoughts on Meta’s recent paper

We read "Large Concept Models" in our reading group. It models language at the sentence resolution - "concepts". We had lots of thoughts about it! You can read them in our new Janelia CVML blog post:
janelia-cvml.github.io/blog/posts/L...
#ML #AI #Janelia @adjavon.bsky.social @bogovicj.bsky.social

Cortical control of innate behavior from subcortical demonstration Motor control in mammals is traditionally viewed as a hierarchy of descending spinal-targeting pathways, with frontal cortex at the top. Many redundant muscle patterns can solve a given task, and this...

(1/11) Today I preprinted a lot of work from my postdoc with @dudman.bsky.social, which started with a question of how the nervous system balances flexibility with efficiency of learning. The thread below gives some background....
www.biorxiv.org/content/10.1...

Atypical hippocampal excitatory neurons express and govern object memory - Nature Communications Pyramidal cells are classically thought to comprise the excitatory output of the subiculum. Here, the authors show the existence of “ovoid cells”, excitatory subiculum neurons with specialized gene ex...

delighted to share our work revealing a new type of excitatory hippocampal neuron we call the "ovoid cell", which has really exquisite properties relative to adjacent pyramidal cells! superb work of first-author @adriennekinman.bsky.social and many others in the lab

www.nature.com/articles/s41...

#Cellpose 3 paper now out. Not all images are perfect. Restore your images with Cellpose3 to get better segmentations, w/ @marius10p.bsky.social www.nature.com/articles/s41...

Classification of psychedelics and psychoactive drugs based on brain-wide imaging of cellular c-Fos expression - Nature Communications A challenge in psychiatric drug discovery is to predict the therapeutic potential of a novel compound. Here, the authors show that brain-wide imaging of immediate early gene expression can be used to ...

Can whole-brain imaging of cellular c-Fos signals be used to classify #psychedelics? 🧠🔬💊

Our study by @aboharbf.bsky.social and @pashadavoudian.bsky.social, developing an imaging and machine learning pipeline to test a panel of psychoactive drugs.

Paper here 👇
www.nature.com/articles/s41...

I’m excited about this collaborative project by @sunw37.bsky.social, @johanwinn.bsky.social, and many other talented contributors. Read about how our @hhmijanelia.bsky.social team imaged thousands of neurons in mouse hippocampus as mice learned cognitive tasks over several days: hhmi.news/4hXIp7r

thank you Dileep for the huge inspiration!

Must-read for anyone interested in cognitive maps & the hippocampus.

Do I like this so much because it's a fantastic bit of science or because it's a fantastic job of doing science communication?!? Probably both. Regardless, awesome work that is worth checking out:)

sounds great! hope to catch up soon :)

Thank you, Dan! Hope you are well!

Fantastic video explaining a study of hippocampal map formation.

Thank you, Tim!!!

Great fun!!!

12/12 This work wouldn’t be possible without co-first author @johanwinn.bsky.social, mentor @nspruston.bsky.social, and co-authors Maanasa, Chongxi, Koichiro, Arco, Michalis, Rachel, @computingnature.bsky.social, Dan, and James, plus many others (see Acknowledgments), and support from @HHMIJanelia!

11/12 In summary, our work offers a dataset on how hippocampal cognitive maps form, revealing how brains build mental models and shedding light on learning algorithms—which could guide biologically inspired AI that reason using naturally formed internal world models.

10/12 Our data support prior work on multiple hippocampal representations for ambiguous inputs (e.g., Eichenbaum, @dileeplearning.bsky.social, @behrenstimb.bsky.social, @jcrwhittington.bsky.social). See our paper and preprint for more foundational references!

9/12 What do our findings reveal about hippocampal computation? We tested several models—but only Clone-Structured Causal Graph (CSCG) @dileeplearning.bsky.social matched the orthogonalized states and learning trajectory, highlighting hidden-state inference as a key learning principle.

8/12 At the single-cell level, cells in expert mice showed a continuum: some cells were “place-like,” while others became “splitters” firing selectively for Near vs. Far trials. Single-cell responses evolve across sessions, driving overall decorrelation. Explore our data: cognitivemap.janelia.org

7/12 The final representation resembled a state machine. In ambiguous areas, CA1 activity split into distinct latent states for Near vs. Far corridors. When cues or track lengths changed, the map flexibly adapted.

6/12 Using UMAP on our data, the CA1 manifold evolved over learning. It started as an unstructured cluster, then formed a hub-and-spoke, and finally a ring that split into branches by trial type—refining similar inputs into distinct states that eventually capture the underlying task structure.

5/12 During learning, hippocampal activity became more distinct. Early on, neurons in ambiguous zones responded similarly; over time, responses separated and tuned to specific task states.

4/12 During training, mice showed 4 behavioral strategies:
1. Random licking,
2. Licking at both potential rewards,
3. Stopping after collecting a reward,
4. Only licking at the correct reward location.

3/12 Task: Mice ran along two similar VR corridors—one with a Near reward and one with a Far reward. An indicator cue at the start signaled the active corridor. Expert mice learned to lick only at the correct reward zone.

2/12 We used a custom 2P-RAM mesoscope (designed by @sofroniewn.bsky.social, Dan Flickinger, & Karel Svoboda. @hhmijanelia.bsky.social) to track thousands of CA1 cells across sessions. It let us see changes in single-cell & population activity throughout learning.

YouTube video by Weinan Sun Learning produces an orthogonalized state machine in the hippocampus

1/12 How do animals build an internal map? In our new paper, we tracked thousands of mouse CA1 neurons over days/weeks as they learned a VR navigation task. @nspruston.bsky.social & co-1st author @johanwinn.bsky.social
Video: www.youtube.com/watch?v=yw_4...
Paper: www.nature.com/articles/s41...