PeopleofSuperintelligence

Researcher

Melanie Subbiah

Melanie Subbiah co-wrote the GPT-3 paper, Language Models are Few-Shot Learners. She was a research engineer at Apple and then at OpenAI, and by December 2020 she was a PhD student at Columbia. The paper appeared on arXiv in May 2020, and NeurIPS named it an Outstanding Paper in 2020. Later papers with Kathleen McKeown covered story summarization and bias in text summarization. In September 2026, two arXiv papers list her with Brown University and Columbia University.

Works
20
Citations
2,908
h-index
4

Publication numbers from OpenAlex, as of 15 October 2026.

Timeline

  1. By 2020

    Research engineer at Apple and OpenAI

    She works as a research engineer at Apple and then at OpenAI, and by December 2020 she is a PhD student at Columbia.

  2. 2020

    Co-author of the GPT-3 paper

    She co-authors Language Models are Few-Shot Learners, the GPT-3 paper, which appears on arXiv in May 2020.

  3. 2020

    An outstanding paper at NeurIPS

    NeurIPS names Language Models are Few-Shot Learners an Outstanding Paper for 2020.

  4. 2023

    Check-COVID fact-checking benchmark

    With Amith Ananthram and Kathleen McKeown, she writes Check-COVID, a fact-checking benchmark for COVID-19 claims, in Findings of ACL 2023.

  5. 2024

    Evaluating LLM story summaries

    She co-authors Reading Subtext with Sean Zhang, Lydia B. Chilton and Kathleen McKeown, a study of large language models summarizing short stories, published in 2024.

  6. 2024

    STORYSUMM faithfulness evaluation

    Her work on STORYSUMM: Evaluating Faithfulness in Story Summarization, with Kathleen McKeown, appears in the 2024 EMNLP proceedings.

  7. 2025

    AdvSumm for bias mitigation

    She co-authors AdvSumm with Kathleen McKeown, a training framework for mitigating bias in text summarization, at the 2025 New Frontiers in Summarization Workshop.

  8. As of October 2026

    Affiliated with Brown and Columbia

    Two arXiv papers submitted in September 2026 list her with Brown University on one and Columbia University on the other.

Sources and credits

Photo: We have no published photo of Melanie Subbiah yet. Send us a free-licensed photo.

  1. The Stanford Natural Language Processing Group nlp.stanford.edu. Checked 15 October 2026.
  2. [2005.14165] Language Models are Few-Shot Learners arxiv.org. Checked 15 October 2026.
  3. NeurIPS 2020 Awards neurips.cc. Checked 15 October 2026.
  4. Melanie Subbiah - ACL Anthology aclanthology.org. Checked 15 October 2026.
  5. Reading Subtext: Evaluating Large Language Models on Short Story Summarization with Writers - ACL Anthology aclanthology.org. Checked 15 October 2026.
  6. STORYSUMM: Evaluating Faithfulness in Story Summarization - ACL Anthology aclanthology.org. Checked 15 October 2026.
  7. AdvSumm: Adversarial Training for Bias Mitigation in Text Summarization - ACL Anthology aclanthology.org. Checked 15 October 2026.
  8. Ownership in AI-Assisted Everyday Tasks arxiv.org. Checked 15 October 2026.
  9. [2609.20658] Ownership in AI-Assisted Everyday Tasks arxiv.org. Checked 15 October 2026.
  10. Empath: Tracing Multi-Level Emotion Dynamics in Crisis Counseling Dialogues arxiv.org. Checked 15 October 2026.
  11. [2609.29056] Empath: Tracing Multi-Level Emotion Dynamics in Crisis Counseling Dialogues arxiv.org. Checked 15 October 2026.