Newest papers

A Neuro-Symbolic Reasoning Engine for Compilation-Free Patch Presence Testing

Siyuan Li, Jiayi Zhou, Zhiyuan Xia, Weijie Wang, Yaowen Zheng, Weiran Tang, Hong Li, Hongsong Zhu · 2026 · arXiv

DOI: 10.48550/arXiv.2610.27871 · open-access PDF

Spanish

La reutilización del código en los ecosistemas de software modernos facilita la propagación de vulnerabilidades de un día, lo que supone riesgos significativos para la seguridad. Determinar si existen parches es crucial, pero los métodos tradicionales sufren de una usabilidad y precisión limitadas, a menudo dependiendo de un proceso de compilación frágil. En nuestro trabajo anterior, introducimos Lares, el primer marco libre de compilación para la prueba de presencia de parches que combina grandes modelos de lenguaje (LLM) con solucionadores SMT. Sin embargo, Lares trata el LLM y el solucionador SMT como dos canales de decisión separados y paralelos: el canal simbólico (Z3) logró sólo el 20% de los resultados.

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

Abstract

Code reuse in modern software ecosystems facilitates the spread of 1-day vulnerabilities, posing significant security risks. Determining if patches are present is crucial, but traditional methods suffer from limited usability and accuracy, often depending on a fragile compilation process. In our prior work, we introduced Lares, the first compilation-free framework for patch presence testing that combines large language models (LLMs) with SMT solvers. However, Lares treats the LLM and the SMT solver as two separate, parallel decision channels: the symbolic channel (Z3) succeeded in only 20% of

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=2610.27871, 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