{
  "id": "benchmark:mmlu",
  "_kind": "Benchmark",
  "_file": "benchmarks/benchmarks/mmlu.yaml",
  "_cluster": "benchmarks",
  "attributes": {
    "displayName": "MMLU",
    "homepageUrl": "https://github.com/hendrycks/test",
    "kind": "knowledge",
    "targetsKind": "ModelVersion",
    "description": "Massive Multitask Language Understanding — 57-subject knowledge benchmark.\n"
  },
  "outgoingEdges": [
    {
      "from": "benchmark:mmlu",
      "to": "skill-area:general-knowledge-reasoning",
      "kind": "covers",
      "attributes": {
        "attributes": {
          "coverage": "full",
          "weight": 0.1
        }
      }
    }
  ],
  "incomingEdges": [
    {
      "from": "eval-result:mmlu.qwen-2-5-72b.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:gpqa.claude-haiku-4-5.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.claude-sonnet-4-6.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.deepseek-v3.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.deepseek-r1.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:gpqa.gemini-2-5-pro.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.llama-4-405b.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.llama-3-1-405b.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.llama-3-3-70b.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.mistral-large-2.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:gpqa.gpt-5.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.o1.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.phi-3-medium.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.gemma-2-27b.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:mmlu.command-r-plus.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-result:gpqa.claude-sonnet-4-5.001",
      "to": "benchmark:mmlu",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-run:mmlu.qwen-2-5-72b.2024-09",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.claude-sonnet-4-6.2025-11",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.deepseek-v3.2024-12",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.deepseek-r1.2025-01",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.llama-4-405b.2024-07",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.llama-3-1-405b.2024-07",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.llama-3-3-70b.2024-12",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.mistral-large-2.2024-07",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.o1.2024-12",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.phi-3-medium.2024-05",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.gemma-2-27b.2024-06",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "eval-run:mmlu.command-r-plus.2024-08",
      "to": "benchmark:mmlu",
      "kind": "for_benchmark"
    },
    {
      "from": "scope-boundary:mmlu.scope",
      "to": "benchmark:mmlu",
      "kind": "bounds_subject"
    }
  ]
}