Bai, Fan
(白帆)
About Me
I build agents at Bloomberg AI. Previously, I was a Postdoctoral Fellow at Johns Hopkins University, where I worked with Prof. Mark Dredze. I received my Ph.D. from Georgia Tech.
Research Interests (in my spare time)
- Agent Safety and Optimization
- Multimodal Learning
- Automated Scientific Discovery
Recent Papers
- Reading, Not Thinking: Understanding and Bridging the Modality Gap When Text Becomes Pixels in Multimodal LLMs
Kaser Sun, Xiaochuang Yuan, Hongjun Liu, Chen Zhao, Cheng Zhang, Mark Dredze, Fan Bai
arXiv 2026
[paper] - FIRE-Bench: Evaluating Agents on the Rediscovery of Scientific Insights
Zhen Wang, Fan Bai, Zhongyan Luo, Jinyan Su, Kaser Sun, Xinle Yu, Jieyuan Liu, Kun Zhou, Claire Cardie, Mark Dredze, Eric P. Xing, Zhiting Hu
arXiv 2026
[blog] [paper] [code] - LLMs are Better Than You Think: Label-Guided In-Context Learning for Named Entity Recognition
Fan Bai, Hamid Hassanzadeh, Ardavan Saeedi, Mark Dredze
EMNLP 2025
[paper] [code] - PromptAgent: Strategic Planning with Language Models Enables Expert-level Prompt Optimization
Xinyuan Wang, Chenxi Li, Zhen Wang, Fan Bai, Haotian Luo, Jiayou Zhang, Nebojsa Jojic, Eric P Xing, Zhiting Hu
ICLR 2024
[paper] [code] - Schema-Driven Information Extraction from Heterogeneous Tables
Fan Bai, Junmo Kang, Gabriel Stanovsky, Dayne Freitag, Alan Ritter
EMNLP 2024 Findings
[paper] [code]