11 papers
Latent On-Policy Self-Distillation
Guibin Zhang, Jiayang Lyu, Ran Sun +4
Enabling agents to learn from experience and internalize it into their policy has become a central problem in self-evolving AI. On-policy self-distillation (OPSD) offers an effecti…
SuperThoughts: Reasoning Tokens in Superposition
Zheyang Xiong, Shivam Garg, Max Yu +4
Long Chain-of-Thought (CoT) reasoning improves LLM problem-solving but is computationally expensive due to sequential token generation. While recent works explore reasoning in cont…
AlgoVeri: An Aligned Benchmark for Verified Code Generation on Classical Algorithms
Haoyu Zhao, Ziran Yang, Jiawei Li +6
Vericoding refers to the generation of formally verified code from rigorous specifications. Recent AI models show promise in vericoding, but a unified methodology for cross-paradig…
Unrealized Expectations: Comparing AI Methods vs Classical Algorithms for Maximum Independent Set
Yikai Wu, Haoyu Zhao, Sanjeev Arora
AI methods, such as generative models and reinforcement learning, have recently been applied to combinatorial optimization (CO) problems, especially NP-hard ones. This paper compar…
Goedel-Code-Prover: Hierarchical Proof Search for Open State-of-the-Art Code Verification
Zenan Li, Ziran Yang, Deyuan He +7
Large language models (LLMs) can generate plausible code but offer limited guarantees of correctness. Formally verifying that implementations satisfy specifications requires constr…
Contextual Drag: How Errors in the Context Affect LLM Reasoning
Yun Cheng, Xingyu Zhu, Haoyu Zhao +1
Central to many self-improvement pipelines for large language models (LLMs) is the assumption that models can improve by reflecting on past mistakes. We study a phenomenon termed c…