Lightnews — Scholar-powered news

Ching Fang

@chingfang.bsky.social

320 followers 260 following 21 posts

Postdoc @Harvard interested in neuro-AI and neurotheory. Previously @columbia, @ucberkeley, and @apple. 🧠🧪🤖

Posts Media Videos Starter Packs

Reposted by Ching Fang

Dionnet L Bhatti Mazo @victoryofflight.bsky.social · Jun 16

Our new preprint: 𝐅𝐞𝐚𝐭𝐮𝐫𝐞-𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐭𝐡𝐫𝐞𝐚𝐭 𝐜𝐨𝐝𝐢𝐧𝐠 𝐢𝐧 𝐥𝐚𝐭𝐞𝐫𝐚𝐥 𝐬𝐞𝐩𝐭𝐮𝐦 𝐠𝐮𝐢𝐝𝐞𝐬 𝐝𝐞𝐟𝐞𝐧𝐬𝐢𝐯𝐞 𝐚𝐜𝐭𝐢𝐨𝐧.

We describe how the LS guides defensive responses by forming critical computations built from functionally and molecularly distinct cells and their afferent inputs.

www.researchsquare.com/article/rs-6...

Feature-specific threat coding in lateral septum guides defensive action

The ability to rapidly detect and evaluate potential threats is essential for survival and requires the integration of sensory information, with internal state and prior experience. The lateral septum...

www.researchsquare.com

Ching Fang @chingfang.bsky.social · Jun 27

Oh cool, thanks for sharing, It does seem like we see very similar things! We should definitely chat 😀

Ching Fang @chingfang.bsky.social · Jun 26

In conclusion: Studying the cognitive computations behind rapid learning requires a broader hypothesis space of planning than standard RL. In both tasks, strategies use intermediate computations cached in memory tokens-- episodic memory itself can be a computational workspace!

Ching Fang @chingfang.bsky.social · Jun 26

In tree mazes, we find a strategy where in-context experience is stitched together to label a critical path from root to goal. If a query state is on this path, an action is chosen to traverse deeper into the tree. If not, the action to go to parent node is optimal. (8/9)

Ching Fang @chingfang.bsky.social · Jun 26

Instead, our analysis of the model in gridworld suggests the following strategy: (1) Use in-context experience to align representations to Euclidean space, (2) Given a query state, calculate the angle in Euclidean space to goal, (3) Use the angle to select an action. (7/9)

Ching Fang @chingfang.bsky.social · Jun 26

Interestingly, when we examine the mechanisms used by the model for decision making, we do not see signatures expected from standard model-free and model-based learning-- the model doesn't use value learning or path planning/state tracking at decision time. (6/9)

Ching Fang @chingfang.bsky.social · Jun 26

We find a few representation learning strategies: (1) in-context structure learning to form a map of the environment and (2) alignment of representations across contexts with the same structure. These connect to computations suggested in hippocampal-entorhinal cortex. (5/9)

Ching Fang @chingfang.bsky.social · Jun 26

As expected, these meta-learned models learn more efficiently in new environments than standard RL since they have useful priors over the task distribution. For instance, models can take shortcut paths in gridworld. So what RL strategies emerged to support this? (4/9)

Ching Fang @chingfang.bsky.social · Jun 26

We train transformers to in-context RL (via decision-pretraining from Lee et al 2023) in planning tasks: gridworld and tree mazes (inspired by labyrinth mazes: elifesciences.org/articles/66175). Importantly, each new task has novel sensory observations. (3/9)

Ching Fang @chingfang.bsky.social · Jun 26

Transformers are a useful setting for studying these questions because they can learn rapidly in-context. But also, key-value architectures have been connected to episodic memory systems in the brain! e.g. see our previous work (of many others) (2/9): elifesciences.org/reviewed-pre...

Barcode activity in a recurrent network model of the hippocampus enables efficient memory binding

elifesciences.org

Ching Fang @chingfang.bsky.social · Jun 26

Humans and animals can rapidly learn in new environments. What computations support this? We study the mechanisms of in-context reinforcement learning in transformers, and propose how episodic memory can support rapid learning. Work w/ @kanakarajanphd.bsky.social : arxiv.org/abs/2506.19686

From memories to maps: Mechanisms of in context reinforcement learning in transformers

Humans and animals show remarkable learning efficiency, adapting to new environments with minimal experience. This capability is not well captured by standard reinforcement learning algorithms that re...

Reposted by Ching Fang

Clementine Domine 🍊 @CCN @clementinedomine.bsky.social · Apr 4

🚀 An other Exciting news! Our paper "From Lazy to Rich: Exact Learning Dynamics in Deep Linear Networks" has been accepted at ICLR 2025!

arxiv.org/abs/2409.14623

A thread on how relative weight initialization shapes learning dynamics in deep networks. 🧵 (1/9)

Reposted by Ching Fang

Jenelle Feather @jfeather.bsky.social · Apr 2

Already feeling #cosyne2025 withdrawal? Apply to the Flatiron Institute Junior Theoretical Neuroscience Workshop! Applications due April 14th

jtnworkshop2025.flatironinstitute.org

Reposted by Ching Fang

Naomi Saphra @nsaphra.bsky.social · Mar 27

Life update: I'm starting as faculty at Boston University
@bucds.bsky.social in 2026! BU has SCHEMES for LM interpretability & analysis, I couldn't be more pumped to join a burgeoning supergroup w/ @najoung.bsky.social @amuuueller.bsky.social. Looking for my first students, so apply and reach out!

CDS building which looks like a jenga tower

Reposted by Ching Fang

Grace Lindsay @neurograce.bsky.social · Mar 26

What are you plans for April 5th? Decide now which event you'll attend and who you'll bring along. See you in the streets!
handsoff2025.com/about

About — Hands Off!

handsoff2025.com

Reposted by Ching Fang

David G. Clark @david-g-clark.bsky.social · Mar 26

I'll be presenting this at #cosyne2025 (poster 3-50)!

I'll also be giving a talk at the "Collectively Emerged Timescales" workshop on this work, plus other projects on emergent dynamics in neural circuits.

Looking forward to seeing everyone in 🇨🇦!

David G. Clark @david-g-clark.bsky.social · Jan 29

(1/30) New preprint! "Symmetries and continuous attractors in disordered neural circuits" with Larry Abbott and Haim Sompolinsky
bioRxiv: www.biorxiv.org/content/10.1...

Symmetries and Continuous Attractors in Disordered Neural Circuits

A major challenge in neuroscience is reconciling idealized theoretical models with complex, heterogeneous experimental data. We address this challenge through the lens of continuous-attractor networks...

www.biorxiv.org

Reposted by Ching Fang

Clementine Domine 🍊 @CCN @clementinedomine.bsky.social · Mar 26

Our paper, “A Theory of Initialization’s Impact on Specialization,” has been accepted to ICLR 2025!
openreview.net/forum?id=RQz...
We shows how neural network can build specialized and shared representation depending on initialization, this has consequences in continual learning.
(1/8)

Ching Fang @chingfang.bsky.social · Mar 24

We'll have a poster on this at #Cosyne2025 in the third poster session (3-055). Come say hi if you're curious!

Ching Fang @chingfang.bsky.social · Mar 24

How does barcode activity in the hippocampus enable precise and flexible memory? How does this relate to key-value memory systems? Our work (w/ Jack Lindsey, Larry Abbott, Dmitriy Aronov, @selmaan.bsky.social ) is now in eLife as a reviewed preprint: elifesciences.org/reviewed-pre...

Barcode activity in a recurrent network model of the hippocampus enables efficient memory binding

elifesciences.org

Ching Fang @chingfang.bsky.social · Mar 24

In particular, barcodes are a plausible neural correlate for the precise slot retrieval mechanism in key-value memory systems (see arxiv.org/abs/2501.02950)! Barcodes provide a content-independent scaffold that binds to memory content, + prevent memories with overlapping content from blurring.

Key-value memory in the brain

Classical models of memory in psychology and neuroscience rely on similarity-based retrieval of stored patterns, where similarity is a function of retrieval cues and the stored patterns. While parsimo...

Ching Fang @chingfang.bsky.social · Mar 24

Why is this useful? We show that place fields + barcode are complementary. Barcodes enable precise recall of cache locations, while place fields enable flexible search for nearby caches. Both are necessary. We also show how barcode memory combines with predictive maps-- check out the paper for more!

Ching Fang @chingfang.bsky.social · Mar 24

A memory of a cache is formed by binding place + seed content to the resulting RNN barcode via Hebbian learning. An animal can recall this memory from place inputs (and high recurrent strength in the RNN). These barcodes capture the spatial correlation profile seen in data.

Ching Fang @chingfang.bsky.social · Mar 24

We suggest a RNN model of barcode memory. The RNN is initialized with random weights and receives place inputs. When recurrent gain is low, RNN units encode place. With high recurrent strength, the random weights produce sparse + uncorrelated barcodes via chaotic dynamics.

Ching Fang @chingfang.bsky.social · Mar 24

We were inspired by @selmaan.bsky.social and Emily Mackevicius' data of neural activity in the hippocampus of food-caching birds during a memory task. Cache events are encoded by barcode activity, which are sparse and uncorrelated patterns. Barcode and place activity coexist in the same population!

Ching Fang @chingfang.bsky.social · Mar 24

How does barcode activity in the hippocampus enable precise and flexible memory? How does this relate to key-value memory systems? Our work (w/ Jack Lindsey, Larry Abbott, Dmitriy Aronov, @selmaan.bsky.social ) is now in eLife as a reviewed preprint: elifesciences.org/reviewed-pre...

Barcode activity in a recurrent network model of the hippocampus enables efficient memory binding

elifesciences.org

Ching Fang @chingfang.bsky.social · Feb 18

We’re organizing a #CoSyNe2025 workshop on what agent models can teach us about neuroscience! See Sat’s thread for more info 😊

Satpreet (Sat) Singh @satpreetsingh.bsky.social · Feb 18

📣 Excited to announce our workshop "Agent-Based Models in Neuroscience: Complex Planning, Embodiment, and Beyond" at the upcoming @cosynemeeting.bsky.social #CoSyNe2025! 🧠🤖 🪱🪰🐟🐝🐭💪

Schedule: neuro-agent-models.github.io
🗓️ Join us in Mont-Tremblant, Canada, on March 31!

neuro-agent-models.github.io