Detecting and Localizing Segment-Level Poisoning in Multi-Source LLM-Agent Inputs

RESUMO

Modern large language model (LLM) agents often construct prompts by aggregating retrieved passages, user reviews, and documents from multiple external sources. This paradigm exposes them to segment-level poisoning attacks, in which an adversary controlling only a small subset of sources injects malicious content to manipulate model outputs. Existing defenses mainly rely on textual patterns, external embeddings, or auxiliary detectors and may therefore fail against fluent, semantically plausible poisoned segments. They also provide limited support for locating the responsible segments. We observe that successful corrupted-evidence and adversarial-instruction attacks induce structured shifts in the LLM's internal activations, forming a consistent activation-space pattern that we call the poison direction. Based on this observation, we propose ActProbe, an internal-state-based framework for detecting and localizing poisoned segments in multi-source LLM inputs. ActProbe projects MLP activations onto a learned poison direction and uses a lightweight linear SVM trained on a small calibration set to detect contaminated prompts. It then applies BinRoL, which combines recursive replacement ablation, Mahalanobis-distance-based branch pruning, and MAD-based robust leaf detection to locate poisoned segments. ActProbe requires no modification to the backend LLM and reduces localization overhead from O(n) exhaustive probing to O(k log n) forward passes. Across three datasets, two attacks, and four open-weight LLMs, ActProbe achieves a 0.01 false-positive rate, a 0.05 false-negative rate, 0.94 localization recall, and a 0.90 localization F1-score. It remains effective against defense-aware adaptive attacks and can protect black-box APIs through surrogate-based poisoned-segment removal.