AI People Wiki

Builders & open source

Mark Chen

Chief Research Officer, OpenAI

Mark Chen is Chief Research Officer at OpenAI, as listed on an OpenAI Forum event page from March 2026. Chen is first author of "Evaluating Large Language Models Trained on Code" (2021), the Codex paper that introduced and released the HumanEval benchmark, and a co-author of the DALL·E and DALL·E 2 papers. The GPT-4 Technical Report credits Chen as vision team co-lead and deployment lead, and the OpenAI o1 System Card lists Chen in the reasoning research leadership and in executive leadership. Chen is also listed among the contributors to the gpt-oss-120b and gpt-oss-20b model card.12345678

At a glance

Current role
Chief Research Officer
Company
OpenAI
Sector
Company
Work listed
5
Sources cited
9
Profile reviewed

Research

Mark Chen is first author of "Generative Pretraining from Pixels" (ICML 2020), which applies a GPT-2-scale sequence transformer to images by predicting pixels autoregressively and measures the resulting representations by linear probing and fine-tuning. Chen is also a co-author of "Zero-Shot Text-to-Image Generation" (2021), the original DALL·E paper, and the last-listed author of the 2022 DALL·E 2 paper on image generation with CLIP latents.945

Chen is first and corresponding author of the 2021 Codex paper, "Evaluating Large Language Models Trained on Code," which introduced HumanEval, a set of programming problems for measuring functional correctness of programs synthesized from docstrings. The openai/human-eval repository, released under the MIT license, provides the evaluation harness and cites the paper.23

Leadership at OpenAI

OpenAI's technical reports credit Chen in several leadership roles. The GPT-4 Technical Report (2023) lists Chen as vision team co-lead and deployment lead, and the OpenAI o1 System Card (2024) lists Chen under reasoning research leadership and executive leadership. Chen is also among the contributors to the 2025 model card for the open-weight gpt-oss-120b and gpt-oss-20b models.678

An OpenAI Forum event page from March 2026 lists Chen as Chief Research Officer at OpenAI and announces a fireside conversation between Chen and mathematician Terence Tao on the role of AI in mathematical discovery, part of a convening with UCLA's Institute for Pure and Applied Mathematics.1

Work and contributions

  • HumanEval Open data

    First author and corresponding author of the Codex paper (2021), which introduced HumanEval, an evaluation set for synthesizing programs from docstrings; the openai/human-eval repository cites the paper and is under the MIT license.23

  • gpt-oss Open weights

    Listed among the contributors to the gpt-oss-120b and gpt-oss-20b model card (2025).8

  • GPT-4 vision and OpenAI o1 Proprietary product

    Credited in the GPT-4 Technical Report (2023) as vision team co-lead and deployment lead, and in the OpenAI o1 System Card (2024) under reasoning research leadership and executive leadership.67

  • DALL·E and DALL·E 2 Published research

    Co-author of "Zero-Shot Text-to-Image Generation" (2021) and last-listed author of "Hierarchical Text-Conditional Image Generation with CLIP Latents" (2022).45

  • Generative Pretraining from Pixels (Image GPT) Published research

    First author of the ICML 2020 paper that trains a sequence transformer to predict pixels autoregressively, without knowledge of 2D image structure, and evaluates the learned image representations.9

Timeline

  1. Published "Generative Pretraining from Pixels" as first author at ICML 2020.9
  2. Co-authored "Zero-Shot Text-to-Image Generation," the DALL·E paper.4
  3. Released the Codex paper and the HumanEval benchmark as first author.23
  4. Co-authored the DALL·E 2 paper on hierarchical text-conditional image generation with CLIP latents.5
  5. Credited as vision team co-lead and deployment lead in the GPT-4 Technical Report.6
  6. Listed under reasoning research leadership and executive leadership in the OpenAI o1 System Card.7
  7. Listed among the contributors to the gpt-oss-120b and gpt-oss-20b model card.8
  8. Listed as OpenAI's Chief Research Officer for a fireside conversation with Terence Tao at an OpenAI and IPAM convening.1

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

9 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. Accelerating Math and Theoretical Physics with AI: OpenAI × UCLA Institute for Pure & Applied Math (speaker listing)OpenAI Forum · Official page · published Mar 4, 2026 · accessed Oct 21, 2026
  2. Evaluating Large Language Models Trained on Code (arXiv 2107.03374)arXiv · Paper · published Jul 7, 2021 · accessed Oct 21, 2026
  3. openai/human-eval repository (README citation; MIT License)OpenAI (GitHub repository) · Repository · accessed Oct 21, 2026
  4. Zero-Shot Text-to-Image Generation (arXiv 2102.12092)arXiv · Paper · published Feb 24, 2021 · accessed Oct 21, 2026
  5. Hierarchical Text-Conditional Image Generation with CLIP Latents (arXiv 2204.06125)arXiv · Paper · published Apr 13, 2022 · accessed Oct 21, 2026
  6. GPT-4 Technical Report (arXiv 2303.08774), authorship and credit attributionarXiv · Paper · published Mar 15, 2023 · accessed Oct 21, 2026
  7. OpenAI o1 System Card (arXiv 2412.16720), authorship and credit attributionarXiv · Paper · published Dec 21, 2024 · accessed Oct 21, 2026
  8. gpt-oss-120b & gpt-oss-20b Model Card (arXiv 2508.10925), contributorsarXiv · Model card · published Aug 8, 2025 · accessed Oct 21, 2026
  9. Generative Pretraining From Pixels (Proceedings of the 37th ICML, PMLR 119)Proceedings of Machine Learning Research · Paper · published 2020 · accessed Oct 21, 2026