SecPO: Principled Adversarial Training for Prompt Injection Security

SecPO introduces a principled adversarial training framework to defend large language models against prompt injection attacks. By approximating an ideal injection‑invariant reference model and employing a new SecPO loss, the authors improve security while preserving utility. Combined with an efficient adversarial training pipeline that synthesizes strong injected inputs, the resulting SecPO + AT model significantly reduces attack success rates compared to existing defenses and maintains performance across multiple benchmarks.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• llms
• agents
• reinforcement-learning
• multimodal
• benchmarks
• safety