works on
agent security 1hidden state analysis 1large language models 1policy violation detection 1prompt injection 1

From the 1 of 1 linked paper with an AI index.

collaborators
Showing cs.CRShow all

1 paper · 1 filter

cs.CR2026

PVDetector: Detecting Prompt Injection Attacks on Purpose-Specific LLM Agents through Policy-Violation Concept Analysis

Junhui Wang, Hangtao Zhang, Zhirun Zheng +5

The paper introduces PVDetector, a training‑free method that detects prompt injection attacks on purpose‑specific LLM agents by measuring alignment of hidden states with policy‑vio…