SIVA: Self-Improving Vulnerability Agent

Valentin Walischewski, Giulio Zizzo, Kevin N. WebsterReliable ML @ NeurIPS 2025 · OpenReview

Abstract

In the ever more digitalized world of today, code vulnerabilities pose a critical threat to our privacy, economy, safety, and infrastructure. Existing automated code vulnerability detection methods suffer from high false positive rates, poor generalization and their inability to adapt to changing vulnerability landscapes. To address these challenges we propose SIVA, a self-improving LLM-based vulnerability detection agent, using memory-guided meta-learning for dynamic prompt optimization. SIVA showed strong learning capabilities, improving its F1 score from 58\% to 95\% in $5$ iterations, significantly outperforming previous state-of-the-art multi-agent systems ($\approx 53\%$ F1) on real-life vulnerability datasets. Furthermore, SIVA generalized well across 7 programming languages (93\% F1), successfully transferring learned vulnerability concepts between them.
BibTeX
@inproceedings{walischewski2025siva,
  author = {Valentin Walischewski and Giulio Zizzo and Kevin N. Webster},
  title = {{SIVA: Self-Improving Vulnerability Agent}},
  booktitle = {Reliable ML @ NeurIPS},
  year = {2025},
  url = {https://openreview.net/forum?id=JDN0x8eTPm},
}

Loading related content...

Figure