[about]   [papers]   [blog]   [travels]   [talks]  


my face


Chenglei Si (チェンレイ・シ)

[email] [scholar] [twitter] [github]

About

 I'm a final-year PhD at the Stanford NLP Group and a researcher at Recursive Superintelligence. At Stanford, I'm advised by the wonderful Tatsu Hashimoto and Diyi Yang. At Recursive, I work on pretraining, autoresearch, and their intersection.

Selected Work
  • Flint Model Release
    core contributor on pretraining
    [blog] [model]

  • First Steps Toward Automated AI Research
    core contributor on autoresearch
    [blog]

  • Towards Execution-Grounded Automated AI Research
    Chenglei Si*, Zitong Yang*, Yejin Choi, Emmanuel Candès, Diyi Yang, Tatsunori Hashimoto
    ICML 2026 [paper] [data] [tweet]

  • The Ideation-Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas
    Chenglei Si, Tatsunori Hashimoto, Diyi Yang
    ICLR 2026 [paper] [data] [tweet] [Science Coverage]

  • Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researchers
    Chenglei Si, Diyi Yang, Tatsunori Hashimoto
    ICLR 2025 [paper] [code] [tweet] [Nature News] [Newsweek] [Synced]

  • Predicting Empirical AI Research Outcomes with Language Models
    Jiaxin Wen, Chenglei Si, Yueh-han Chen, He He, Shi Feng
    NeurIPS 2025 [paper] [tweet]

  • Design2Code: How Far Are We From Automating Front-End Engineering?
    Chenglei Si*, Yanzhe Zhang*, Zhengyuan Yang, Ruibo Liu, Diyi Yang
    NAACL 2025 [paper] [code] [project page] [tweet] [WorldofAI]

Talks

Conference Travels
  • Oct 2026, COLM @ San Francisco (co-organizer, LM4Sci 2.0)
  • Apr 2026, ICLR @ Rio de Janeiro
  • Oct 2025, COLM @ Montreal (co-organizer, LM4Sci)
  • Apr 2025, ICLR @ Singapore
  • Dec 2024, NeurIPS @ Vancouver
  • Jun 2024, NAACL @ Mexico City
  • Oct 2023, UIST @ San Francisco
  • July 2023, ACL @ Toronto
  • May 2023, ICLR @ Kigali
  • Dec 2022, EMNLP @ Abu Dhabi
  • July 2022, NAACL @ Seattle