About

I am a fifth year CS Ph.D. candidate at The University of North Carolina, Chapel Hill advised by Prof. Mohit Bansal. My research focuses on multimodal AI and agentic systems for understanding and reasoning across different modalities (especially videos). My goal is to develop effective and efficient systems that integrate perception and reasoning to solve complex real-world problems.

I am on the job market, looking for full-time Research Scientist positions starting Summer 2027. Feel free to reach out!

Research Interests

In general, I am interested in building multimodal systems that perceive and reason over long and complex videos. My recent work includes agentic video understanding (Searching Videos as Trees, NeurIPS 2026; Active Video Perception, CVPR 2026; VideoTree, CVPR 2025), long-horizon memory and reasoning (EgoMemReason, COLM 2026), evidence-grounded and verifiable reasoning (Evidence-Backed VideoQA, ECCV 2026; Multimodal Fact-Level Attribution, ICML 2026), RL + test-time scaling (Video-RTS, EMNLP 2025), and LLM-based video understanding (LLoVi, MEXA, SiLVR). Earlier, I worked on video-text retrieval (UCoFiA, ICCV 2023) and zero-shot learning.

NewsRecent

• 2026Searching Videos as Trees accepted to NeurIPS 2026.
• 2026Evidence-Backed Video Question Answering accepted to ECCV 2026.
• 2026EgoMemReason accepted to COLM 2026.
• May 2026Joined Bloomberg in New York as a Data Science Fellow and Research Intern, working on large-scale multimodal document understanding.
• 2026Multimodal Fact-Level Attribution for Verifiable Reasoning accepted to ICML 2026.
• Mar 2026Received the Bloomberg Data Science Ph.D. Fellowship.
• 2026Active Video Perception accepted to CVPR 2026 Findings.
• 2026Co-organizing the Transformers for Vision (T4V) Workshop at CVPR 2026.
• 2026SiLVR accepted to TMLR.
• 2026TimeRefine accepted to WACV 2026.
• 2025Video-RTS accepted to EMNLP 2025 Main and MEXA to EMNLP 2025 Findings.
• May 2025Joined Salesforce AI Research in Palo Alto as a Research Intern, working on agentic long video understanding.
• 2025VideoTree accepted to CVPR 2025.
• 2025Co-organizing the Transformers for Vision (T4V) Workshop at CVPR 2025.
• 2025DAM accepted to WACV 2025.
• 2024LLoVi accepted to EMNLP 2024 Main.
• May 2024Joined the Meta FAIR Perception team in Menlo Park as a Research Intern.
• 2024Unified Embeddings for Multimodal Retrieval via Frozen LLMs accepted to EACL 2024 Findings.
• 2023UCoFiA accepted to ICCV 2023.
• May 2023Joined Amazon Alexa AI in Seattle as an Applied Scientist Intern.
• Aug 2022Started my Ph.D. at UNC-Chapel Hill, advised by Prof. Mohit Bansal.

Publications

By Topic By Year

Experience

• May 2026 — Aug 2026
  Data Science Fellow and Research Intern · Bloomberg

  New York City, NY · Large-scale multimodal document understanding, with Dr. Debanjan Mahata and Dr. Daniel Preoțiuc-Pietro
• May 2025 — Feb 2026
  Research Intern · Salesforce AI Research

  Palo Alto, CA · Agentic long video understanding, with Dr. Juan Carlos Niebles, Dr. Michael S. Ryoo, Dr. Junnan Li and Dr. Honglu Zhou
• May 2024 — Dec 2024
  Research Intern · Meta FAIR Perception

  Menlo Park, CA · Video foundation encoders, with Dr. Ronghang Hu, Dr. Christoph Feichtenhofer, Dr. Po-Yao Huang and Dr. Daniel Bolya
• May 2023 — Oct 2023
  Applied Scientist Intern · Amazon Alexa AI

  Seattle, WA · Multimodal retrieval with LLMs, with Dr. Heba Elfardy, Dr. Kevin Small and Dr. Markus Dreyer
• Oct 2021 — May 2022
  Research Intern · Tsinghua AIR

  Beijing, China · Advised by Prof. Jingjing Liu

Education

• 2022 — present
  Ph.D. in Computer Science

  UNC-Chapel Hill · Advised by Prof. Mohit Bansal
• 2022 — 2024
  M.S. in Computer Science

  UNC-Chapel Hill · During Ph.D. study
• 2018 — 2022
  B.Eng. in Software Engineering

  University of Electronic Science and Technology of China · Advised by Prof. Jingjing Li

Honors & Service

I am a die-heart Arsenal and Tar Heel fan.