II.
Benchmark JSON
Structured · livebenchmark:m-mmlu
Multilingual MMLU (mMMLU) json
Inspect the normalized record payload exactly as the atlas UI reads it.
{
"id": "benchmark:m-mmlu",
"_kind": "Benchmark",
"_file": "benchmarks/benchmarks/benchmarks-knowledge.yaml",
"_cluster": "benchmarks",
"attributes": {
"displayName": "Multilingual MMLU (mMMLU)",
"homepageUrl": "https://huggingface.co/datasets/openai/MMMLU",
"kind": "model-only",
"targetsKind": "ModelVersion",
"description": "A multilingual translation of MMLU (Massive Multitask Language\nUnderstanding) covering 14 languages, used for evaluating cross-\nlingual knowledge and reasoning. Released by OpenAI as MMMLU.\n"
},
"outgoingEdges": [
{
"from": "benchmark:m-mmlu",
"to": "skill-area:general-knowledge-reasoning",
"kind": "covers",
"attributes": {}
},
{
"from": "benchmark:m-mmlu",
"to": "domain:education",
"kind": "applies_to",
"attributes": {}
}
],
"incomingEdges": []
}