He Cheng
  • Bio
  • Papers
  • Projects
  • Talks
  • News
  • Services
  • Experience
  • CV

Projects

Research projects in biomedical NLP, knowledge graphs, multi-agent systems, and anomaly detection.

My research spans efficient knowledge graph retrieval, multi-agent clinical reasoning, and interpretable and robust anomaly detection.

Orchestrator-Centered Multi-Agent Systems for Clinical Reasoning

Specialized large language model agents generate, review, and refine diagnostic predictions through structured multi-step interactions, coordinated by a reinforcement-learning orchestrator.

Multi-Agent SystemsLLMsReinforcement LearningClinical AI

BadSAD

Clean-label backdoor attacks against deep semi-supervised anomaly detection models.

Official record · Preprint

Anomaly DetectionBackdoor AttacksTrustworthy AI

LogosKG

Hardware-optimized, scalable, and interpretable multi-hop retrieval over large knowledge graphs.

Paper · Code · Demo

Knowledge GraphsLLMsBiomedical NLP

BLOG

A backdoor attack framework designed for log anomaly detection models.

Publication

Anomaly DetectionBackdoor AttacksLog Data

CFDet

Counterfactual explanations for sequence anomaly detection that identify minimal changes needed to convert an anomalous sequence into a normal one.

Publication

Anomaly DetectionCounterfactual ExplanationSequence Modeling

BA-OCAD

A clean-label backdoor attack framework targeting one-class sequential anomaly detection models.

Publication

Anomaly DetectionBackdoor AttacksSecurity

EASD

Prototype-based explainable sequential anomaly detection with human-understandable explanations.

Publication

Anomaly DetectionExplainabilityPrototypes

GLEAD

Sequential anomaly detection with both local explanations and global explanations.

Publication

Anomaly DetectionInterpretabilitySequential Data

 

© 2026 He Cheng