MMLU: Philosophy

Accuracy on MMLU's philosophy questions, one of 57 subject tests of academic and professional knowledge.

Also known as: philosophy

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.
Categoryknowledge
SubcategoryHumanities
Page statusactive
Metricaccuracy
Directionhigher_is_better
Unit%
Dataset size311
Dataset licenceMIT
PublisherUC Berkeley

What it measures

A general philosophy survey spanning epistemology, metaphysics, ethics, and logic, and the major positions and figures associated with each. Framed as four-option multiple-choice questions and scored zero-shot or few-shot by exact match against the labelled option, as one of the 57 subject subsets that make up the MMLU benchmark.

Task format

Four-option multiple-choice question answering (A-D), one correct answer, zero-shot or few-shot.

Models reporting this benchmark

These figures come from the model cards, which carry one collection date per card and no per-score attribution. They are shown as reported, not as verified evidence.
ModelProviderScoreCard as of
Mixtral 8x22B Instruct v0.1Mistral AI82.62026-04
Yi 34B 200K01.AI82.02024-07
Meta Llama 3 70BMeta81.42024-07
Meta Llama 3 70B InstructMeta81.42026-04
Meta Llama 3 70B InstructNous Research81.42026-04
Nous Hermes 2 Yi 34BNous Research81.42024-07
Yi 1.5 34B01.AI81.42026-04
Yi 34B Chat01.AI81.02024-07
Yi 1.5 34B 32K01.AI80.42024-07
Mixtral 8x7B Instruct v0.1Mistral AI79.72026-04
Nous Hermes 2 Mixtral 8x7B DPONous Research79.12024-07
Mixtral 8x7B v0.1Mistral AI78.52026-04
Yi 1.5 34B Chat01.AI77.52026-04
Yi 1.5 34B Chat 16K01.AI77.52024-07
Yi 9B01.AI77.52024-07
Yi 1.5 9B 32K01.AI76.82024-07
Phi 3 mini 4K instructMicrosoft76.22024-07
Yi 1.5 9B Chat 16K01.AI75.92024-07
Yi 1.5 9B01.AI75.62024-07
Meta Llama 3 8BMeta74.32024-07
Meta Llama 3 8BNous Research74.32024-07
Hermes 2 Theta Llama 3 8BNous Research73.62024-07
Phi 3 mini 128K instructMicrosoft73.62024-07
Yi 1.5 9B Chat01.AI73.62024-07
Mistral 7B v0.3Mistral AI72.72024-07
mistral 7B v0.3 bnb 4bitUnsloth72.72024-07
Nous Hermes 2 SOLAR 10.7BNous Research72.72024-07
gemma 7B itGoogle DeepMind72.32024-07
Meta Llama 3 8B InstructMeta71.42024-07
Meta Llama 3 8B InstructNous Research71.42024-07
Hermes 2 Pro Llama 3 8BNous Research71.12024-07
Mistral 7B Instruct v0.2Mistral AI70.42024-07
Yi 6B01.AI69.12024-07
Yi 6B Chat01.AI69.12024-07
Yi 1.5 6B01.AI68.52024-07
falcon 40BTII66.92024-07
Qwen2 1.5B InstructAlibaba / Qwen Team64.32024-07
Yi 1.5 6B Chat01.AI64.02024-07
phi 2Microsoft62.12024-07
deepseek llm 7B baseDeepSeek55.62024-07
deepseek llm 7B chatDeepSeek55.62024-07
Qwen2 0.5B InstructAlibaba / Qwen Team50.52024-07
chatglm2 6BZhipu AI49.82024-07
deepseek coder 6.7B baseDeepSeek44.12024-07
deepseek coder 6.7B instructDeepSeek42.82024-07
gemma 2BGoogle DeepMind41.22024-07
gemma 2B itGoogle DeepMind40.22024-07
deepseek coder 1.3B baseDeepSeek29.32024-07
deepseek coder 1.3B instructDeepSeek29.32024-07
OLMo 1B hfAllen AI28.92024-07

Data

This page as JSON · Edit on GitHub