I am a PhD candidate at Stony Brook University, advised by Owen Rambow.

I work on evaluation, post-training, multi-modality, and human-AI interaction.

My research connects LLMs to everyday tasks, such as personalized writing assessment & imitation (ACL 2025, EMNLP 2025), and multi-turn (interactive) visual grounding (EMNLP 2025, ACL 2026, EMNLP 2026).

In industry, I have worked on bringing (multimodal) LLMs to customer-facing applications, including topic modeling, feed ranking, speech understanding (via SFT), creative ad image editing (via RL), and search/recommendation.