Shen Zheng

Hi! I am a Senior Research Scientist at Bytedance Seed Team, working on code LLMs and Agents. I got my MS in CS from UIUC and decided to move into industry rather than PhD. Before that, I obtained my bachelor degree in Zhejiang University.
My research work includes:
- Automated curation for scaling posttrain data
- Test-time scaling
- LLM pretraining
- LLM in coding and math [1] [2]
Previously in academia, I worked on:
selected publications & models
- Model
- Model
- NAACLGPT-Fathom: Benchmarking Large Language Models to Decipher the Evolutionary Path towards GPT-4 and BeyondFindings of the Association for Computational Linguistics: NAACL 2024, 2024
- ACLBFS-Prover: Scalable Best-First Tree Search for LLM-based Automatic Theorem ProvingACL 2025, 2025