AI RESEARCH
Detecting Fluent Optimization-Based Adversarial Prompts via Sequential Entropy Changes
arXiv CS.AI
•
ArXi:2605.19966v1 Announce Type: cross Optimization-based adversarial suffixes can jailbreak aligned large language models (LLMs) while remaining fluent, weakening static and windowed perplexity-based detectors. We cast adversarial suffix detection as an online change-point detection problem over the token-level next-token entropy stream. Using the LLM system prompt to estimate a robust baseline, we standardize user-token entropies and apply a one-sided CUSUM statistic. The resulting detector, CPD Online (CPD), is model-agnostic.