I'm Melanie Subbiah.
I am a postdoctoral researcher in Computer Science, working with Ellie Pavlick at Brown University. I work on interpretation and interpretability with language models in subjective domains, often with a focus on human narrative. In a past life, I worked on GPT-3 at OpenAI. I will be on the job market for research faculty positions this year.
Education
Columbia University
Ph.D. in Computer Science
Thesis: Interpretation and Interpretability in Subjective Domains
Advisor: Kathleen McKeown
Awards:
- PhD Departmental Service Award
- Best Paper Award for TACL 2024 (Reading Subtext: Evaluating Large Language Models on Short Story Summarization with Writers)
- Amazon CAIT PhD Fellowship 2023
- NSF GRFP Honorable Mention 2021 (Comp/IS/Eng - Natural Language Processing)
- Outstanding Paper Award at NeurIPS 2020 (Language Models are Few-Shot Learners)
- Presidential and SEAS Fellowship 2020
Columbia University
M.S. in Computer Science
Completed as part of MS/PhD
Williams College
B.A. in Computer Science, with highest honors
Graduated magna cum laude
Thesis: Using Text Abstraction and LSTM Language Models for Domain-Independent Narrative Generation
Advisor: Andrea Danyluk
Awards:
- Phi Beta Kappa Commencement Student Speaker
- Best Thesis Presentation award in Computer Science
- Magna Cum Laude, Phi Beta Kappa, Sigma Xi
- Honorable Mention award in short story writing
Work
Meta
Research Intern
Neural interfaces research - deep learning with EMG data.
OpenAI
Member of Technical Staff
Generative language models research - GPT-3 and the OpenAI API.
Apple
Machine Learning Research Engineer
Autonomous systems research - robotics, RL, computer vision, and NLP.
Research
H Mian*, M Subbiah*, S Marcus, N Shaalan, K McKeown. Computational Representations of Character Significance in Novels. In submission. 2026. > PDF
M Subbiah, A Mishra, G Kim, L Tang, G Durrett, K McKeown. Is the Top Still Spinning? Evaluating Subjectivity in Narrative Understanding. EMNLP. 2025. > PDF
M Limpijankit, Y Chen, M Subbiah, N Deas, K McKeown. Counterfactual Simulatability of LLM Explanations for Generation Tasks. INLG. 2025. > PDF
M Gupt, N Varimall, N Deas, M Subbiah, K McKeown. AdvSumm: Adversarial Training for Bias Mitigation in Text Summarization. NewSumm Workshop at EMNLP. 2025. > PDF
Z Hall, M Subbiah*, T Zollo, K McKeown, R Zemel. Guiding LLM Decision-Making with Fairness Reward Models. NeurIPS. 2025.
A Mayukha, M Subbiah, A Guzman, S Jitklongsub, D McAdams. Methodological Approaches to the Analysis of Life Story Interviews. ARP. 2025. > PDF
M Subbiah, F Ladhak, A Mishra, G Adams, LB Chilton, K McKeown. StorySumm: Evaluating Faithfulness in Story Summarization. EMNLP. 2024. > PDF
Best Paper Award
M Subbiah, S Zhang, LB Chilton, K McKeown. Reading Subtext: Evaluating Large Language Models on Short Story Summarization with Writers. TACL. 2024.
> PDF
A Storek, M Subbiah, K McKeown. Unsupervised Selective Rationalization with Noise Injection. ACL. 2023. > PDF
G Wang, L Chillrud, K Harwood, A Ananthram, M Subbiah, K McKeown. Check-COVID: Fact-Checking COVID-19 News Claims with Scientific Evidence. Findings of ACL. 2023. > PDF
M Subbiah*, A Bhattacharjee*, Y Hua, TS Kumarage, H Liu, K McKeown. Detecting Harmful Agendas in News Articles. WASSA Workshop at ACL. 2023. > PDF
Red Teaming for GPT-4. 2023. > PDF
S Levy, E Allaway, M Subbiah, L Chilton, D Patton, K McKeown, WY Wang. SafeText: A Benchmark for Exploring Physical Safety in Language Models. EMNLP. 2022. > PDF
A Mei, A Kabir, S Levy, M Subbiah, E Allaway, J Judge, D Patton, B Bimber, K McKeown, WY Wang. Mitigating Covertly Unsafe Text within Natural Language Systems. Findings of EMNLP. 2022. > PDF
M Subbiah, K McKeown. Understanding Identity Signalling in Persuasive Online Text. International Workshop on Social Sensing at ICWSM. 2021. > PDF
Co-organizer of ICLR Workshop on Enormous Language Models. 2021. > workshop website
Outstanding Paper Award
TB Brown*, B Mann*, N Ryder*, M Subbiah*, et al. Language Models are Few-Shot Learners. NeurIPS. 2020.
> PDF
M Subbiah, et al. Augmenting Training Data with Simulated Images. Women in Machine Learning Workshop at NeurIPS. 2018.
M Maher, M Subbiah, N Apostoloff. Cascaded Dataset QA. Women in Machine Learning Workshop at NeurIPS. 2018.
UC Nygaard, Z Li, T Palys, B Jackson, M Subbiah, M Malipatlolla, V Sampath, H Maecker, MR Karagas, KC Nadeau. Cord blood T cell subpopulations and associations with maternal cadmium and arsenic exposures. PLoS One. > PDF
Speaking
Invited Computer Science colloquium talk at Williams College. 2024. > Event
Academy for Teachers' Master Class: What Is ChatGPT? How Do You Use It?. 2023. > Event
Wall Street Journal's The Journal: Artifial - Episode 2, Selling Out. 2023. > Podcast
PhD Candidacy Exam: Narrative Summarization. 2023. > Slides
Pearson Publishing's AI Catalyst Conference: NLP with ChatGPT. 2023. > Event
SuperDataScience podcast: GPT-3 for Natural Language Processing. 2022. > Podcast
Wired 5 Levels of Difficulty video: Machine Learning. YouTube. 2021. > Video
Invited seminar talk on GPT-3 at Stanford University. 2020. > Event
Invited seminar talk on GPT-3 at New York University. 2020. > Slides
Invited Computer Science colloquium talk at Williams College. 2020. > Event
Williams College Commencement Phi Beta Kappa speaker. 2017. > Video
Teaching
- Guest lecture for Computing in Context at Columbia
- Guest lecture for Global Teaching Labs at MIT - Uruguay
- Full instructor for Discrete Mathematics at Columbia
- Guest lecture for Computational Journalism at Columbia
- TA (and lecture) for Natural Language Generation and Summarization at Columbia
- CS Tutor for Columbia Athletics
- TA for Intro CS and Data Structures at Williams
Mentorship
- Columbia University research students - 3 undergraduate, and 7 masters
- Williams College CS undergrad mentor
- OpenAI Scholars program development
- OpenAI Scholars mentor
- Institute of International Education TechWomen mentor
- Williams College Underrepresented Identities in CS leader
Interests
- Natural Language Processing
- Language Models
- Narrative Understanding
- Reinforcement Learning
- Deep Learning
- Computational Creativity
- AI for Climate
- AI for Energy Efficiency
Profiles
Contact
I enjoy creating AI systems that are capable of interpreting and reasoning about human behavior and social dynamics. I value ethical and sustainable approaches to AI research.