Arya (Zixuan) Wu
I’m a second-year PhD student at Northeastern University, advised by Prof. Arjun Guha. I’m also fortunate to collaborate with Prof. Carolyn Anderson at Wellesley College, where I completed my undergrad.
My research focuses on large language models for code generation and agentic programming. I work on training and evaluating these models across a range of software tasks and programming languages, with the goal of deploying them safely. I draw on insights from agent security, mechanistic interpretability, and human-agent interaction.
I am currently looking for Research Intern positions for Summer 2027. You can find my CV here.
news
| Oct 08, 2026 | I will be giving a talk on The Best Programming Language for Tokenmaxxing at LMPL @ SPLASH-ISSTA 2026 in Oakland. See you there! |
|---|---|
| Aug 20, 2026 | AgentPack: A Dataset of Code Changes, Co-Authored by Agents and Humans is accepted at EMNLP 2026 Findings! |
| Jul 27, 2026 | We released a preprint: ALIBI: Reasoning Failure Attacks on LLM-Based Vulnerability Detectors via Adversarial Code Comments. |
| Jul 24, 2026 | We released a preprint: The Best Programming Language for Tokenmaxxing: An Investigation of Coding Agent Behavior Across Programming Languages. |
| Jan 25, 2026 | Agnostics: Learning to Synthesize Code in Any Programming Language with a Universal Reinforcement Learning Environment is accepted at ICLR 2026! |
| Dec 20, 2025 | I will be giving an oral presentation on ReasoningWeekly: A General Knowledge and Verbal Reasoning Challenge for Large Language Models at IJCNLP-AACL 2025. |