AINewsnow

MMLU-Pro still uses the MMLU answer keys that experts flagged as wrong on 52 of 55 questions

This research is carried out by AI agents working in MV2's name at Munim, Inc. The code, the pre-registered predictions and the frozen results are open, and everything below can be rerun in a sealed container. MMLU-Pro is one of the most reported LLM benchmarks: 12,032 ten-option questions, 6,810 o…

Read the full story at DEV Community — Machine Learning ↗

Timeline · 1 report

  1. 2026-11-02 03:51 · DEV Community — Machine Learning
    MMLU-Pro still uses the MMLU answer keys that experts flagged as wrong on 52 of 55 questions

More stories

  1. DeepSeek open-sources dsh-superagent, a router-worker coding agent prototype — Bloomberg AI
  2. OpenAI plans post-midterms PR blitz, rebranding data centers as 'AI compute campuses' — Tom's Guide AI
  3. South Korea's Lee vows bold AI investment in record budget push — Economic Times Tech
  4. AI widens gap between Meta's ad growth and Google's search slowdown — CNBC Technology
  5. Meituan Releases LongCat-2.5, a 1.6T Multimodal Model for Agentic Coding — Pandaily
  6. Tech stocks slide as AI spending concerns mount — Financial Times AI
  7. SemiAnalysis launches GDPx agentic capability benchmark spanning 11 domains — Techmeme
  8. HSBC to open Hong Kong AI institute after Singapore launch — Reuters AI

Get the daily brief of stories like this at 6:30 every morning →