paper-with-me

홈 › Papers

Automating Formal Verification with Reinforcement Learning and Recursive Inference

2026-05-29 · Max Tan arxiv

Automated formal verification remains challenging for large language models because data for proof assistants and verification-aware languages is scarce, and correctness depends on satisfying precise machine-checkable specifications rather than producing plausible code. This thesis studies how verifier environments can improve LLM generation of verified programs and proofs through reinforcement learning from verifiable rewards (RLVR) and verifier-guided inference-time search. First, we train open-source models in Dafny with RLVR using Group Relative Policy Optimization (GRPO) and related variants, assembling generated candidates into complete programs and scoring them with compiler and verifier outcomes. Initial experiments on an APPS-derived Dafny dataset increased verified reward from 2.2% to 58.1%, but revealed specification hacking, where models exploit weak formal specifications instead of implementing the intended solutions. After filtering underspecified and vulnerable tasks, multi-turn RLVR on the refined benchmark improves the verified pass rate from 9.7% to 31.1%. Second, we develop a verifier-guided inference scaffold in Lean that treats proof generation as structured search over decomposed subgoals, verifier feedback, diagnostics, and repair. With a fixed base model, the full scaffold with proof reviser improves pass rate on an initial VeriCoding pilot set from 46.2% under direct repair to 69.2%. On the larger VERINA dataset, whole-task decomposition plus proof reviser solves 7 of 42 previously unsolved tasks. We also introduce Dalek-Bench, a repository-scale Lean benchmark derived from the Rust $\texttt{curve25519-dalek}$ verification project; preliminary results remain weak, indicating that stronger progress evaluation and task-specific tool-use policies are still needed.

📄 PDF Abstract BibTeX arXiv:2605.30914

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

QEDCartographer: Automating Formal Verification Using Reward-Free Reinforcement Learning

2024-08-17 · Alex Sanchez-Stern, Abhishek Varghese, Zhanna Kaufman, Dylan Zhang 외

Formal verification is a promising method for producing reliable software, but the difficulty of manually writing verification proofs severely limits its utility in practice. Recent methods have automated some proof synt…

reinforcement-learningReinforcement Learning

From Scientific Texts to Verifiable Code: Automating the Process with Transformers

2025-01-09 · Changjie Wang, Mariano Scazzariello, Marco Chiesa

Despite the vast body of research literature proposing algorithms with formal guarantees, the amount of verifiable code in today's systems remains minimal. This discrepancy stems from the inherent difficulty of verifying…

Decompose-and-Formalise: Recursively Verifiable Natural Language Inference

2026-01-27 · Xin Quan, Marco Valentino, Louise A. Dennis, André Freitas arxiv

Recent work has shown that integrating large language models (LLMs) with theorem provers (TPs) in neuro-symbolic pipelines helps with entailment verification and proof-guided refinement of explanations for natural langua…

Natural Language Inference

Tool Verification for Test-Time Reinforcement Learning

2026-03-02 · Ruotong Liao, Nikolai Röhrich, Xiaohan Wang, Yuhui Zhang 외 arxiv

Test-time reinforcement learning (TTRL) has emerged as a promising paradigm for Recursive Self-Improving AI (RSI) by adapting Large Reasoning Models (LRMs) on unlabeled test inputs, using self-consensus rewards derived f…

Reinforcement Learning

TerraFormer: Automated Infrastructure-as-Code with LLMs Fine-Tuned via Policy-Guided Verifier Feedback

2026-01-13 · Prithwish Jana, Sam Davidson, Bhavana Bhasker, Andrey Kan 외 arxiv

Automating Infrastructure-as-Code (IaC) is challenging, and large language models (LLMs) often produce incorrect configurations from natural language (NL). We present TerraFormer, a neuro-symbolic framework for IaC gener…

Reinforcement Learning