Newest papers

VeruSteer: Learning What to Repair Next in Verus

Chao Jiang, Ding Wang, Chaohua Yang, Dugang Liu, Zhiwu Xu, Cheng Wen, Zhong Ming, Yuwei Liu · 2026 · arXiv

DOI: 10.48550/arXiv.2610.29755 · open-access PDF

Portuguese

Sistemas de verificação baseados em LLM reparam provas de Rust revisando programas em resposta ao diagnóstico de Verus. Quando vários tipos de erros são relatados, pipelines como o AutoVerus e o VeruSAGE selecionam o próximo alvo de reparação usando uma prioridade pré-estabelecida. No entanto, cada reparação altera o programa e os erros que se seguem, de modo que os mesmos tipos de negócios podem exigir escolhas diferentes em estados diferentes. Nós apresentamos o VeruSteer, que aprende a selecionar tipos de erro a partir dos resultados de trajetórias de reparação alternativas. Ele constrói árvores de reparação offline e compara caminhos do mesmo programa inicial por sucesso de verificação

Machine translation (Meta M2M100); the English original follows.

Abstract

LLM-based verification systems repair Rust proofs by revising programs in response to Verus diagnostics. When several error types are reported, pipelines such as AutoVerus and VeruSAGE select the next repair target using a preset priority. Yet each repair changes the program and the errors that follow, so the same reported types can require different choices at different states. We present VeruSteer, which learns to select error types from the outcomes of alternative repair trajectories. It constructs repair trees offline and compares paths from the same initial program by verification success

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.29755, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange