Tag: Evaluation

1. AI agents now finish 16 percent of real freelance jobs, up from 2.5 (safe.ai)
   AI · 20 hours ago · July 15, 2026
2. Why price per million tokens tells you almost nothing (janilowski.pl)
   AI · 1 week ago · July 7, 2026
3. GeneBench-Pro shows AI still stumbles on real biology analysis (openai.com)
   AI · 1 week ago · July 5, 2026
4. Using DSPy to find a hidden bug in an agent's prompt (simonwillison.net)
   AI · 1 week ago · July 3, 2026
5. A new benchmark shows agents rebuilding software that would take humans weeks (epoch.ai)
   AI · 2 weeks ago · June 30, 2026
6. An open model beat Claude Code at finding access-control bugs (semgrep.dev)
   Security · 2 weeks ago · June 29, 2026
7. An API change that helps big models can break small ones (huggingface.co)
   AI · 3 weeks ago · June 21, 2026
8. GLM-5.2 takes the lead among open-weights models (artificialanalysis.ai)
   AI · 3 weeks ago · June 18, 2026
9. OpenAI predicts model behavior by replaying real conversations (openai.com)
   AI · 3 weeks ago · June 18, 2026
10. AI2's olmo-eval brings statistical rigor to the model training loop (huggingface.co)
    AI · 1 month ago · June 13, 2026
11. Did Claude make rsync buggier? The numbers say no (alexispurslane.github.io)
    AI · 1 month ago · June 5, 2026
12. Microsoft ASSERT turns plain-English behavior specs into AI tests (techcrunch.com)
    Engineering · 1 month ago · June 4, 2026
13. Princeton researchers pick apart Google's claim that AI agents built an OS for $916 (normaltech.ai)
    AI · 1 month ago · May 30, 2026
14. A new benchmark sends frontier agents to fix Kubernetes, and none pass (huggingface.co)
    AI · 1 month ago · May 27, 2026
15. An open leaderboard for whole agent systems, not just models (huggingface.co)
    AI · 1 month ago · May 19, 2026
16. Ai2 launches a shared benchmark for AI climate models (allenai.org)
    AI · 1 month ago · May 18, 2026
17. The Open ASR Leaderboard adds private tests to stop gaming (huggingface.co)
    AI · 1 month ago · May 18, 2026
18. Testing AI in the open world, not just on benchmarks (normaltech.ai)
    AI · 1 month ago · May 17, 2026
19. How you pick benchmarks decides whether open models are far behind (interconnects.ai)
    AI · 2 months ago · May 16, 2026
20. One benchmark number hides which jobs a model is actually good at (interconnects.ai)
    AI · 2 months ago · April 20, 2026
21. DeepMind proposes a cognitive framework for measuring AGI progress (blog.google)
    AI · 4 months ago · March 17, 2026
22. Crash Testing GPT-4: The First Dangerous-Capability Eval (asteriskmag.com)
    AI · 3 years ago · June 1, 2023

← all tags