Zeyneb N. Kaya
Hi! I am Zeyneb, I study CS+Math at Stanford University. I'm broadly interested in understanding and pushing the limits of ML, working on robustness/adaptation, learning from data (efficiently), statistics, and physics––among other things.
Recently, projects I've worked on include reasoning/IOL @ OpenAI, models for physics+optimization as co-founder @ Topological; decentralized AI, synthetic data, and midtraining @ Dria; and RL/self-improvement/diffusion @ SAIL.
I’m always eager to discuss interesting ideas—please reach out! When I'm not reading papers, I'll geek out over poetry/art, geoguessr/linguistics, my collections, cats, andwhatever topic I've spiraled into.
zeynebnk [at] stanford [dot] edu
Research.
My interests are in studying and building systems that learn, with work broadly spanning self-improvement/continual learning, data/robustness, and learning dynamics, working in machine learning and statistics.
Listed below are selected relevant research.
Anchored Self-Play for Code Repair
Caroline Choi, Zeyneb N. Kaya, Shirley Wu, Tengyu Ma, Tatsunori Hashimoto, Ludwig Schmidt
preprint, 2026
Test-Time Meta-Adaptation with Self-Synthesis
Zeyneb N. Kaya, Nick Rui
ICLR RSI, DATA-FM, & LIT Workshops, 2026
Optimizing Remasking Schedules for Reasoning in Discrete Diffusion Models
Zeyneb N. Kaya, Radostin Cholakov, Nicole H. Ma
ICLR ReALM-GEN & NFAM Workshops, 2026
Semantic Anchoring in Large Language Models: Thresholds, Transfer, and Geometry
Edward Y. Chang, Ethan Chang, Zeyneb N. Kaya
arXiv preprint, 2026
Measuring the Impact of Data Augmentation Methods for Extremely Low-Resource NMT
Zeyneb N. Kaya, Annie K. Lamar
EACL LoResMT Workshop @ EACL, 2023
MADLIBS: A Novel Multilingual Data Augmentation Algorithm for Low-Resource Neural Machine Translation
Zeyneb N. Kaya
Regeneron Science Talent Search, 2024
Full Scope Word Embedding Variability for Low-Resource Languages
Zeyneb N. Kaya, Annie K. Lamar
IEEE MIT URTC, 2023
Zeyneb N. Kaya
Proceedings of the Linguistic Society of America (PLSA), 2023
Zeyneb N. Kaya, Souvick Ghosh
arXiv preprint
Awards & Recognition.
Etched x Mercor x Cognition Hackathon – 1st Place/$40K Winner 2025
Regeneron Science Talent Search Winner – 5th Place/$90K Winner 2024
Coca Cola Scholar – 2024
PearVC x Anthropic Hackathon – 1st Place/Most Technical Winner, 2025
TreeHacks Scrapybara Prize – 1st Place/$16K Winner, 2025
Geoguessr – Master Tier Player, 2025
National Junior Science and Humanities Symposium (NJSHS) – National HM, 2nd Math/CS, 2023
Congressional App Challenge – 1st Place Winner, 2021
Olympiad in Linguistics (Onling) – 10th Place / 1st in USA, 2023
International Olympiad in Artificial Intelligence – Team USA invited representative (did not attend due to conflicts)
Projects & Blogs.
to be included
Spin Glasses and the Statistical Mechanics of Transformers
Topological and Algebraic Connectivity in Random Graphs
Spiky Smooth Shapes in High Dimensions
Language Models (can be) Few-Shot Fakers
The Shape of Data & Learning
