Technology ❯ Artificial Intelligence ❯ Machine Learning

Benchmarking

Performance Evaluation Performance Metrics MLPerf Evaluation Metrics ARC-AGI Humanity’s Last Exam Grok 4 vs GPT-5 Quality Metrics GPT-5 Multi-Modal Benchmarking

ARC Prize Releases ARC-AGI-3, Harder Agent Benchmark With $1 Million Prize

The new suite shifts focus to experience-driven learning using replayable runs with tooling that makes agent behavior inspectable.

Moonshot Releases Open-Source Kimi K2 Thinking, Claiming Wins Over GPT-5 on Reasoning Tests

Moonshot AI Releases Open-Source Kimi K2 Thinking, a Tool-Using 'Thinking Agent'

New Data Shows AI Is Reshaping Work More Than Cutting Jobs

Musk Says Grok 5 Training Starts Within Weeks, Claims Shot at AGI

Developers Embrace RAG to Ground Language Models in Accurate, Up-to-Date Data

Meta Faces Backlash Over Use of Experimental AI Model for Benchmark Testing

Nvidia's Blackwell GPUs Set New AI Training Benchmarks

FrontierMath Benchmark Reveals AI's Struggles with Complex Math

Nvidia Confirms Blackwell GPU Production Despite Design Defect

Nvidia Dominates AI Benchmarks with Record-Breaking Performance