6 papers · 1 filter
Verifier-Grounded Agentic RL for Provably Correct Code
George Ma, Yuhao Zhang, Bing He +5
Reinforcement learning for code usually rewards passing tests, which a model can do while remaining wrong on the inputs the tests do not cover. A program verifier instead decides w…
SWE-Proof: Can Language Models Resolve Real-World Issues with Machine-Checked Proofs?
George Ma, Benjamin Mikek, Haoyu Li +9
Ensuring the correctness of LLM-generated code is a core challenge for modern software engineering. Benchmarks for agentic code generation check correctness with held-out test suit…
Do Sparse Autoencoders Identify Reasoning Features in Language Models?
George Ma, Zhongyuan Liang, Irene Y. Chen +1
We study how reliably sparse autoencoders (SAEs) support claims about reasoning-related internal features in large language models. We first give a stylized analysis showing that s…
Spooky Action at a Distance: Normalization Layers Enable Side-Channel Spatial Communication
Samuel Pfrommer, George Ma, Yixiao Huang +1
This work shows that normalization layers can facilitate a surprising degree of communication across the spatial dimensions of an input tensor. We study a toy localization task wit…
Baking Symmetry into GFlowNets
George Ma, Emmanuel Bengio, Yoshua Bengio +1
GFlowNets have exhibited promising performance in generating diverse candidates with high rewards. These networks generate objects incrementally and aim to learn a policy that assi…
A Canonicalization Perspective on Invariant and Equivariant Learning
George Ma, Yifei Wang, Derek Lim +2
In many applications, we desire neural networks to exhibit invariance or equivariance to certain groups due to symmetries inherent in the data. Recently, frame-averaging methods em…