MMLU: Prehistory

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

Also known as: prehistory

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 size324
Dataset licenceMIT
PublisherUC Berkeley

What it measures

Human prehistory and archaeology before written records: early hominids, stone-age technology, and the methods archaeologists use to study this period. 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
Meta Llama 3 70BMeta90.12024-07
Meta Llama 3 70B InstructMeta90.12026-04
Meta Llama 3 70B InstructNous Research90.12026-04
Nous Hermes 2 Yi 34BNous Research88.92024-07
Mixtral 8x22B Instruct v0.1Mistral AI87.02026-04
Yi 34B Chat01.AI87.02024-07
Yi 1.5 34B 32K01.AI86.12024-07
Nous Hermes 2 Mixtral 8x7B DPONous Research85.52024-07
Yi 1.5 34B Chat01.AI85.22026-04
Yi 1.5 34B Chat 16K01.AI85.22024-07
Yi 1.5 34B01.AI84.62026-04
Mixtral 8x7B v0.1Mistral AI84.02026-04
Yi 34B 200K01.AI83.62024-07
Mixtral 8x7B Instruct v0.1Mistral AI83.32026-04
Phi 3 mini 4K instructMicrosoft78.12024-07
Nous Hermes 2 SOLAR 10.7BNous Research77.52024-07
Phi 3 mini 128K instructMicrosoft77.22024-07
Yi 9B01.AI76.92024-07
Yi 1.5 9B01.AI75.32024-07
Yi 1.5 9B 32K01.AI74.42024-07
Meta Llama 3 8B InstructMeta74.12024-07
Meta Llama 3 8B InstructNous Research74.12024-07
Yi 1.5 9B Chat 16K01.AI74.12024-07
gemma 7B itGoogle DeepMind73.82024-07
Hermes 2 Theta Llama 3 8BNous Research73.82024-07
Yi 1.5 9B Chat01.AI73.12024-07
Meta Llama 3 8BMeta72.52024-07
Meta Llama 3 8BNous Research72.52024-07
Hermes 2 Pro Llama 3 8BNous Research71.32024-07
Yi 1.5 6B01.AI71.32024-07
Yi 6B01.AI71.02024-07
Yi 6B Chat01.AI71.02024-07
Mistral 7B Instruct v0.2Mistral AI70.12024-07
Mistral 7B v0.3Mistral AI68.82024-07
mistral 7B v0.3 bnb 4bitUnsloth68.82024-07
Yi 1.5 6B Chat01.AI63.62024-07
falcon 40BTII63.32024-07
phi 2Microsoft62.02024-07
Qwen2 1.5B InstructAlibaba / Qwen Team59.02024-07
deepseek llm 7B baseDeepSeek56.52024-07
deepseek llm 7B chatDeepSeek56.52024-07
chatglm2 6BZhipu AI49.42024-07
gemma 2BGoogle DeepMind46.32024-07
Qwen2 0.5B InstructAlibaba / Qwen Team43.82024-07
gemma 2B itGoogle DeepMind40.72024-07
deepseek coder 6.7B instructDeepSeek36.42024-07
deepseek coder 1.3B baseDeepSeek28.72024-07
deepseek coder 1.3B instructDeepSeek28.72024-07
deepseek coder 6.7B baseDeepSeek28.12024-07
OLMo 1B hfAllen AI25.62024-07

Data

This page as JSON · Edit on GitHub