SOURCE-LINKED INTELLIGENCE

SelfOp: An Optimization Algorithm for Self-Improving Security Agents

arXiv · AI, language, vision and robotics · article · Sep 19, 2026 · UTC

LLM agents are increasingly used for security tasks: vulnerability discovery, exploit reproduction, and patch generation. Improving them at the model level demands expert demonstrations or computable rewards, which security tasks rarely offer: traces are costly, failures hard to diagnose, rewards sparse, and non-computable. Efforts thus shift to the harness and context, but manual tuning needs task-specific expertise and scales poorly, while automated methods rely on scarce ground truth, stronger optimizer models, or unguided propose-and-evaluate loops that reduce to costly trial and error. We

Read original source ↗ Open in workspace

recordType

paper

region

Global