2 papers
cs.SE2026
Accepted but Wrong: Silent Failures of LLM Code Repairs after API Migrations
Xiaobo Chen, Genliang Zhu, Ziyuan Wang +1
After a runtime or API migration, an LLM-repaired procedure can run cleanly yet return wrong results, and a release check on typical test inputs ships it. We report a preregistered…
cs.CR2026
Signed but Not Safe: Requirement Integrity in Multi-Agent LLM Handoffs
Xiaobo Chen, Genliang Zhu, Ziyuan Wang +1
Enterprise AI agents increasingly split a user's request across subagents and merge their results, and each handoff can summarize away, alter or ignore a requirement. We report a p…