Haixing He

works on

defense mechanisms 1 information bottleneck 1 intent-aware detection 1 LLM agents 1 memory injection attacks 1

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

collaborators

1 paper

cs.AI2026

MIND: Lightweight and Effective Memory Injection Defense for LLM Agents via Intent-Aware Information Bottleneck

Dongyi Liu, Haixing He, Xiaobao Wu +1

The paper introduces MIND, a lightweight framework that uses an intent‑aware information bottleneck to detect and filter poisoned memory in large language model agents, reducing at…