Research · Researcher

Allen Liu

“Assistant Professor of CS at NYU Courant | ML Foundations”

LinkedIn X GitHub Website

Joined: Jun 2026
How they joined: Directly
Before Cognition: NYU Courant Institute School of Mathematics, Computing, and Data ScienceAssistant Professor
Olympiads: IMO gold 2014–2016IMO 2014GoldIMO 2015GoldIMO 2016Gold

Overlapped before Cognition

Sources¹cphof.org ²linkedin.com ³linkedin.com

Autumn profile

Researched with Autumn · 2026-11-04

Details

LocationUnited States

LanguagesChinese (Simplified), English

UniversityMassachusetts Institute of Technology

ResearchC++PythonMachine LearningMathematicsMachine Learning AlgorithmsTheoretical Computer ScienceLearning TheoryAlgorithms

Notes

Listed among the contributors on Cognition's SWE-1.7 launch post (2026-07-08), roughly four weeks after joining; SWE-1.7 was RL-trained on a Kimi K2.7 base and shipped in Devin Web, Desktop and CLI at 1000 TPS via Cerebras.Jul 8, 2026
Cognition publicly announced that Allen joined its research team the week of 2026-06-11 "while continuing as a professor at NYU", so the Cognition role is concurrent with, not a replacement for, his NYU Courant faculty post.Jun 11, 2026
Cognition's announcement framed the hire around his ACM Doctoral Dissertation Award for work on "learning-theoretic foundations for understanding quantum systems"; colleagues Alex Kaplan replied "Honored to work with you @axliu42!" and Jay Grant "huge win for the research team", identifying two Cognition-side warm contacts.Jun 11, 2026
PhD (2025) and SM (2022) in CS at MIT advised by Ankur Moitra, BS in Mathematics from MIT (2020); then Miller postdoctoral fellow at UC Berkeley in fall 2025 hosted by Prasad Raghavendra, and NYU Courant assistant professor from January 2026.Jan 2026
Recent LLM theory with Noah Golowich and Abhishek Shetty, e.g. 'Sequences of Logits Reveal the Low Rank Structure of Language Models' (ICLR 2026 oral) and 'Model Stealing for Any Low-Rank Language Model' (STOC 2025, with Moitra).2026
Three-time IMO gold medalist (2014-2016) with a perfect score at IMO 2016, per his CV.2016
His only public X activity about Cognition is reposts of @cognition's SWE-1.7 launch threads (2026-07-08) and the FrontierCode eval launch (2026-06-09); he has not written original posts about working at Cognition.Jul 8, 2026
Rohan Choudhury posted "so excited to work with you @axliu42 !!! @cognition", another Cognition research colleague who publicly welcomed him.Jun 11, 2026
Core quantum-learning collaboration with Ainesh Bakshi, Ankur Moitra and Ewin Tang produced 'Learning Quantum Hamiltonians at any Temperature in Polynomial Time' (QIP 2024 best student paper, STOC 2024) and 'High-Temperature Gibbs States are Unentangled' (FOCS 2024), both covered by Quanta.2024
Research internships at Google Research NY (2019 hosted by Jon Schneider and Renato Paes Leme; 2020 hosted by Morteza Zadimoghaddam) and Microsoft Research WA (2021, 2022 hosted by Jerry Li); Jerry Li later became a frequent coauthor on quantum learning papers.2022
Before academia he did a trading internship at Jane Street Capital (NY, summer 2017) and a quantitative research internship at D. E. Shaw & Co. (NY, summer 2018).2017
Public headline says "Assistant Professor of CS at NYU Courant | ML Foundations".

Autumn profile

People they worked with before

More from Research