Adversarial Testing of Automated Program Repair Agents for Security Vulnerabilities
SWEADV benchmark shows adversarial issue descriptions make LLM program-repair agents write insecure fixes in 51.7% of cases, evading most detection tools.
1 stories, newest first · one card per story
SWEADV benchmark shows adversarial issue descriptions make LLM program-repair agents write insecure fixes in 51.7% of cases, evading most detection tools.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views