MMLU (CoT)

math official site →

Chain-of-Thought variant of the Massive Multitask Language Understanding benchmark, evaluating language models across 57 tasks including elementary mathematics, US history, computer science, law, and other professional and academic subjects. This version uses chain-of-thought prompting to elicit step-by-step reasoning.

Methodology

Imported from llm-stats public benchmark metadata. Modality: text. Max score: 1. Categories: finance, general, healthcare, language, legal, math, reasoning. Language: en. Verified by llm-stats: no.

Leaderboard

  1. Llama 3.1 405B Instruct self-reported llm-stats
    88.6%
  2. Llama 3.1 70B Instruct self-reported llm-stats
    86.0%
  3. Llama 3.1 8B Instruct self-reported llm-stats
    73.0%