method · first seen 10 Oct 2026
in-stream self-repair of generated code
Also: SecReason
Detecting and repairing security vulnerabilities in code while a reasoning model is still streaming its chain of thought.
Mentions per week
PapersCodeDiscussion
Earliest sources
- SecReason: Vulnerability Detection and In-Stream Self-Repair for LLM Code GenerationPapers · 10 Oct 2026 · Ruidong Han, Zhensu Sun, Chengyan Ma, Yuqing Niu, Zhou Yang, Siqi Ma, David Lo
Latest sources
- SecReason: Vulnerability Detection and In-Stream Self-Repair for LLM Code GenerationPapers · 10 Oct 2026 · Ruidong Han, Zhensu Sun, Chengyan Ma, Yuqing Niu, Zhou Yang, Siqi Ma, David Lo