MMLU: World Religions

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

Also known as: world_religions

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 size171
Dataset licenceMIT
PublisherUC Berkeley

What it measures

Comparative religion: the beliefs, practices, and history of major world religions, including Christianity, Islam, Judaism, Hinduism, and Buddhism. 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 AI90.62026-04
Yi 1.5 34B 32K01.AI90.62024-07
Meta Llama 3 70BMeta90.12024-07
Meta Llama 3 70B InstructMeta90.12026-04
Meta Llama 3 70B InstructNous Research90.12026-04
Nous Hermes 2 Mixtral 8x7B DPONous Research89.52024-07
Yi 1.5 34B01.AI88.92026-04
Mixtral 8x7B Instruct v0.1Mistral AI87.72026-04
Mixtral 8x7B v0.1Mistral AI87.72026-04
Yi 34B 200K01.AI87.72024-07
Yi 34B Chat01.AI87.72024-07
Nous Hermes 2 Yi 34BNous Research87.12024-07
Yi 1.5 34B Chat01.AI86.52026-04
Yi 1.5 34B Chat 16K01.AI86.52024-07
Yi 1.5 9B Chat 16K01.AI84.22024-07
Mistral 7B Instruct v0.2Mistral AI83.62024-07
Phi 3 mini 4K instructMicrosoft83.62024-07
Yi 1.5 9B 32K01.AI83.62024-07
Meta Llama 3 8BMeta83.02024-07
Meta Llama 3 8BNous Research83.02024-07
Nous Hermes 2 SOLAR 10.7BNous Research83.02024-07
Yi 1.5 9B01.AI83.02024-07
Yi 6B01.AI83.02024-07
Yi 6B Chat01.AI83.02024-07
Yi 9B01.AI83.02024-07
Yi 1.5 9B Chat01.AI82.52024-07
gemma 7B itGoogle DeepMind81.92024-07
Hermes 2 Pro Llama 3 8BNous Research81.92024-07
Mistral 7B v0.3Mistral AI81.92024-07
mistral 7B v0.3 bnb 4bitUnsloth81.92024-07
falcon 40BTII80.72024-07
Hermes 2 Theta Llama 3 8BNous Research80.72024-07
Phi 3 mini 128K instructMicrosoft80.72024-07
Meta Llama 3 8B InstructMeta77.82024-07
Meta Llama 3 8B InstructNous Research77.82024-07
deepseek llm 7B baseDeepSeek76.02024-07
deepseek llm 7B chatDeepSeek76.02024-07
Yi 1.5 6B01.AI76.02024-07
Qwen2 1.5B InstructAlibaba / Qwen Team74.32024-07
Yi 1.5 6B Chat01.AI74.32024-07
phi 2Microsoft69.62024-07
gemma 2BGoogle DeepMind54.42024-07
Qwen2 0.5B InstructAlibaba / Qwen Team48.52024-07
chatglm2 6BZhipu AI47.42024-07
gemma 2B itGoogle DeepMind43.92024-07
deepseek coder 6.7B baseDeepSeek38.62024-07
deepseek coder 6.7B instructDeepSeek35.12024-07
OLMo 1B hfAllen AI25.72024-07
deepseek coder 1.3B baseDeepSeek21.62024-07
deepseek coder 1.3B instructDeepSeek21.62024-07

Data

This page as JSON · Edit on GitHub