Ph.D. in computer science, focus on XAI, LLM safety and reasoning, Agents
rethink_sft_generalization. Repo for paper "Rethinking Generalization in Reasoning SFT: A Conditional Analysis on Optimization, Data, and Model Capability"
bottleneck. PyTorch implementation of the paper "Discovering and Explaining the Representation Bottleneck of DNNs" (ICLR 2022 Oral)
