LLM 智能体跨客户目标泄漏：KV cache 中的目标表征如何影响决策

View PDF HTML (experimental)

Abstract:As LLM agents take on more autonomous roles, they may interact with multiple clients and counterparties whose goals and incentives conflict. If an agent maintains shared inference state across these interactions, information associated with one relationship may influence decisions made for another. We test OLMo-3-7B-Instruct in a controlled two-client environment with 20 market scenarios, where each client is assigned an opposing objective. With shared conversational memory, the model deviates from the current client's objective in 22% of eligible decisions. We then investigate how these goals are represented in the KV cache. Linear probes show that client-goal ownership becomes highly recoverable in middle layers and is concentrated in particular attention heads. We use these results to identify where to intervene in the cache and test whether these representations affect the model's decisions. Ablating the competitor's cached goal representations at these locations shifts the model toward the current client. The effect is larger for heads identified by probing than for randomly selected heads, and larger when ablating goal tokens than matched non-goal tokens from the same client. Our results show that client-specific goal representations persist in the KV cache and can causally influence decisions made for other clients, highlighting the importance of relationship-specific state isolation in persistent LLM agents.

Submission history

From: Saloni Gupta [view email]
[v1] Mon, 12 Oct 2026 20:22:45 UTC (3,987 KB)