I am a first-year M.S. student in Computer Science at Brown University, working on natural language processing. My research focuses on understanding and controlling large language models (LLMs). I am broadly interested in interpretability, model control, and reinforcement learning, with the goal of developing more reliable and trustworthy language models and agents.
My work explores how insights into model representations and decision-making can inform more effective post-training and inference-time control. This includes methods for interpreting and steering language models, as well as reinforcement learning approaches for improving the behavior of language-model agents. Ultimately, I aim to build models and agents that generalize reliably, behave safely, and remain controllable in complex environments.
I am currently a research intern at UCLA NLP in Summer 2026. I also work closely with Prof. Kuan-Hao Huang at Texas A&M University. Previously, I received my Bachelor's degree in Computer Engineering with First Class Honors from Nanyang Technological University, where I was fortunate to be advised by Prof. Wenya Wang. I am happy to collaborate on research in model control, agentic RL, and other exciting problems of NLP and AI.
We propose In-Context Routing, an implicit in-context learning method that captures reusable ICL routing structure and modulates attention logits to simulate few-shot behavior without explicit demonstrations at inference time.
We study structure-aware exemplar selection for in-context learning in complex semantic parsing tasks, and introduce a structural alignment perspective for selecting more effective demonstrations.
We introduce In-Context Prompt Tuning, an efficient framework for personalizing large vision-language models by transforming fine-grained visual semantics from reference images into adaptive continuous prompts.
We introduce TRACES, a proactive safety auditing framework that models the evolving trajectory state of multi-turn LLM agents to detect unsafe behavior before task completion.
We introduce Steering Vector Fields, a context-aware framework for inference-time control that adapts steering directions according to local representation geometry.