TRACE: A Multi-Layer Benchmark for Human AI Controller Coordination Under Drift and Failure
Joshua Zuniga, Srinivasan Subramanian, Ramya Madhuri Narapureddy, Md Abdullah Al Hafiz Khan
cs.AI, cs.HC
Submitted: 2026-08-07
Updated: 2026-08-10
Comments: This work was accepted for presentation and publication at the 25th IEEE International Conference on Machine Learning and Applications (ICMLA 2026)
Code: https://github.com/m0nkspade/TRACE-benchmark
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Bayesian Online Changepoint Detection
- Why Do Multi-Agent LLM Systems Fail?
- ALFRED: A Benchmark for Interpreting Grounded Instructions for Everyday Tasks
- ReAct: Synergizing Reasoning and Acting in Language Models
- Attention Is All You Need
- A comprehensive analysis of concept drift locality in data streams
- TrajAD: Trajectory Anomaly Detection for Trustworthy LLM Agents
- Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems