Ching Fang (chingfang.bsky.social)
@chingfang17
theoretical neuroscience postdoc at @Harvard. Previously @Columbia @cu_neurotheory, @Apple, @UCBerkeley.
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: arxiv.org/abs/2506.19686
We're launching an "AI psychiatry" team as part of interpretability efforts at Anthropic! We'll be researching phenomena like model personas, motivations, and situational awareness, and how they lead to spooky/unhinged behaviors. We're hiring - join us! job-boards.greenhouse.io/anthropic/jobs…
🧠 Can a neural network build a spatial map from scattered episodic experiences like humans do? We introduce the Episodic Spatial World Model (ESWM)—a model that constructs flexible internal world models from sparse, disjoint memories. 🧵👇 [1/12]
I'll be presenting this work today at #NeurIPS2024 medical foundation models workshop! Stop by if you're curious
Sharing my internship work on multimodal foundation models for physiol. signals (EEG, EMG, EOG, ECG)! This is work with the Body-sensing Intelligence Group @Apple that will be presented @NeurIPSConf Medical Foundation Models (AIM-FM) workshop. arxiv.org/abs/2410.16424
Excited to share our new work on dynamics in compositional RL with curricula accepted @icmlconf 2024! 📄 [arxiv.org/abs/2402.18361] 👯 Team = me, @stefsmlab @SaxeLab TL;DR We propose a simple compositional RL model amenable to exact learning dynamics analysis which can account…