HBEE: Human Behavioral Entropy Engine -- Pre-Registered Multi-Agent LLM Simulation of Peer-Suspicion-Based Detection Inversion
On the map: LLM agents and tool use → Agent threats, guardrails and governance
Abstract (arXiv)
Insider threat detection assumes that an adaptive insider leaves behavioral residue distinguishing them from legitimate users. We test this assumption against an LLM-driven adaptive insider in a controlled multi-agent simulator. Our pre-registered five-condition study isolates defender mode (cascade vs. blind UEBA) crossed with adversary type (naive vs. adaptive OPSEC) plus a no-mole control, across 100 runs (95 valid after pre-committed exclusions). The primary finding is a detection inversion: at T_60, the adaptive mole's suspicion in-degree is statistically lower than a randomly selected innocent agent (Cliff's delta = -0.694, 95% BCa CI [-0.855, -0.519], Mann-Whitney p << 0.01). The pre-registered prediction was the opposite direction. A pre-registered equivalence test (H2) shows adaptive OPSEC produces no detectable shift in the mole's UEBA rank under either defender mode. The two detection signals (peer suspicion graph in-degree and per-agent UEBA rank) decouple under adaptive adversary behavior. We bound generalization explicitly: a pre-registered Gini calibration check (H4) returns FAIL, with HBEE pairwise message-exposure Gini (0.213) diverging from the SNAP Enron reference (0.730) by |Delta Gini| = 0.52, exceeding the equivalence bound by 5x. The paper makes a narrow but surprising claim: in a controlled environment where adaptive OPSEC is implementable as an LLM directive, peer-suspicion-cascade detection inverts. We release the simulator, pre-registration document, frozen scenarios, raw telemetry, and analysis pipeline under an open-source license.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use. Builds on 10 of its 24 references that are in the map. First posted on arXiv 8 May 2026.
Builds on
10 of 24 references.
Closest ancestors · 2
Same direction, newest first
- 2025Multimodal Safety Evaluation in Generative Agent Social Simulations9 Oct
- 2025Chimera: Harnessing Multi-Agent LLMs for Automatic Insider Threat Simulation11 Aug
Other references · 8
- 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
- 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
- 2021GLM: General Language Model Pretraining with Autoregressive Blank Infillingcited by 1,026
- 2020Improving Reproducibility in Machine Learning Research (A Report from the NeurIPS 2019 Reproducibility Program)cited by 318
- 2025AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Societycited by 180
- 2023Simulating Social Media Using Large Language Models to Evaluate Alternative News Feed Algorithmscited by 64
- 2020Deep Learning for Insider Threat Detection: Review, Challenges and Opportunitiescited by 18
- 2026OrgForge-IT: A Verifiable Synthetic Benchmark for LLM-Based Insider Threat Detectioncited by 4
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2025Who's the Mole? Modeling and Detecting Intention-Hiding Malicious Agents in LLM-Based Multi-Agent Systems
- 2026Early Detection of Distributed Backdoors in Multi-Agent LLM Systems: A Characterization Study
- 2026Honeyquest for LLMs: Rethinking Cyber Deception for AI Attackers
- 2026Adaptive Adversaries: A Multi-Turn, Multi-LLM Benchmark for LLM Agent Security
- 2024Audit-LLM: Multi-Agent Collaboration for Log-based Insider Threat Detection
- 2026Adaptive Evaluation of Out-of-Band Defenses Against Prompt Injection in LLM Agents
- 2026Agent-ToM: Learning to Monitor Autonomous LLM Agents via Theory-of-Mind Reasoning
- 2026When Collaboration Becomes a Trigger: Collective Evidence-Threshold Backdoors in Multi-Agent Systems