Tag: AI evaluation
Browse our exclusive articles!
Predicting LLM Agent Performance via Cumulative State Tracking
Discover how cumulative state tracking with WMF-AM improves prediction of LLM agent performance beyond task completion rates.
Uncertainty-Aware XAI: A Systematic Survey & Insights
Explore a comprehensive survey on uncertainty-aware XAI, covering methods, evaluation challenges, and future directions for reliable AI explanations.
AIDABench: Benchmarking AI Data Analytics Performance
Discover AIDABench, a comprehensive AI benchmark for evaluating data analytics across QA, visualization, and file generation tasks in real-world scenarios.
Understanding Chain-of-Thought Divergence in AI Reasoning Models
Explore how open-weight reasoning models reveal thinking-answer divergence and its impact on AI transparency and reliability in reasoning tasks.
Stress-Test Benchmark for Multi-Subject Identity Preservation
Explore a new stress-test benchmark revealing identity collapse in multi-subject AI models and the innovative SCR metric for better evaluation.