LATIDIA · Ciberseguridad

Ataques de envenenamiento sigilosos a través de la caché de razonamiento KV reutilizada

arXiv:2610.22873v1 Announce Type: new Resumen: Los sistemas modernos de modelo de lenguaje grande (LLM) reutilizan cachés de clave-valor (KV) para acelerar la inferencia. Trabajos recientes han demostrado que la reutilización de cachés KV a partir de rastros de razonamiento también puede i

La noticia

arXiv:2610.22873v1 Announce Type: new Abstract: Modern large language model (LLM) systems reuse key-value (KV) caches to accelerate inference. Recent work has shown that reusing KV caches from reasoning traces can also improve task performance, similar to retrieval-augmented generation (RAG). We demonstrate that this form of KV cache reuse introduces new poisoning vulnerabilities. We propose R$^2$ECAP (Reused Reasoning-Cache Poisoning), a black-box attack that plants poisoned reasoning traces solely through interactions with a system's chat interface. R$^2$ECAP constructs two traces: a preceding trace that steers the model toward an adversarial target and a retained trace that carries this target. Because the retained trace is generated conditioned on the preceding trace, its KV