Robust Agent Compensation: Enhancing AI Agent Reliability
Discover Robust Agent Compensation (RAC), a log-based recovery framework that boosts AI agent reliability and performance without code changes.
GeoDecider: Explainable Coarse-to-Fine Lithology Classification
Discover GeoDecider, a multi-stage workflow enhancing accuracy and explainability in lithology classification with expert-like geological reasoning.
ReasonAudio: Benchmark for Advanced Text-Audio Reasoning
Discover ReasonAudio, a benchmark designed to evaluate advanced reasoning in text-audio retrieval beyond simple matching tasks.
Inside Agent Memory: Circuit Analysis & Failure Diagnosis
Explore how agent memory circuits in LLMs emerge, operate, and fail, revealing key insights for improving AI reliability and diagnostics.
LLM-Powered Automated Solver for Large-Scale CVRP
Discover how LLM-assisted Flexible MCTS automates large-scale CVRP solver design, enhancing efficiency and scalability in vehicle routing solutions.