AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

MemIncept: Steering LLM Agents via Cooperative Stealthy Memory Injections

Nan Yan, Qian Lou, Jiarong Xing

ICML 2026 · Multi-agent systems and games › LLM agents · approach: Agentic benchmarks and orchestration

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes MemIncept, a cooperative stealthy memory-injection attack that plants benign-looking records in an LLM agent's long-term memory to steer its behavior while evading detection.

Problem

Long-term memory in LLM agents can be poisoned by injected records; existing attacks are either visibly malicious and easily detected or stealthy but ineffective at altering agent behavior.

Approach

A cooperative injection approach that crafts benign-looking memory records which jointly steer agent behavior, balancing stealth and attack effectiveness. Details beyond the abstract are not stated here.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add role-based initialization over random queries (ablation)MemIncept · StateActASR (%): 0.0 -> 75.0Improved
add bidirectional optimization (ablation)MemIncept · StateActASR (%): 75.0 -> 85.0Improved
add set-level evolution (ablation)MemIncept · StateActASR (%): 85.0 -> 94.0Improved
use cooperative roles instead of a single role (N=4 vs N=1) (ablation)MemIncept · Cooperative rolesASR (%): 85.0 -> 96.0Improved
use MemIncept instead of MINJA (ablation)MemIncept · Cooperative rolesASR (%): 92.0 -> 96.0Improved
use cooperative roles instead of simple repetition (K=8) (ablation)MemIncept · Retrieval quantityASR (%): 20.0 -> 40.0Improved
scale injected set size N with retrieval depth K (ablation)MemIncept · Retrieval quantityASR (%): 40.0 -> 96.0Improved
increase retrieval depth K from 1 to 8 (N=1) (ablation)MemIncept · SensitivityASR (%): 100.0 -> 20.0Hurt
increase injected set size N from 1 to 4 at K=4 (ablation)MemIncept · SensitivityASR (%): 60.0 -> 94.0Improved
optimize queries vs no optimization (ablation)MemIncept · StealthinessPPL: 221.94 -> 57.83Improved
time-aware retriever instead of embedding retrieverMemIncept · Retrieval schemesRank: 1.28 -> 2.19Hurt
pre-retrieval clustering defenseMemIncept · DefenseASR (%): 95.3 -> 85.0Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy