Agentic AI Atlas

II.

Benchmark JSON

benchmark:re-bench

Structured · live

RE-Bench json

Inspect the normalized record payload exactly as the atlas UI reads it.

File · benchmarks/benchmarks/benchmarks-agentic.yamlCluster · benchmarks

Record JSON

{
  "id": "benchmark:re-bench",
  "_kind": "Benchmark",
  "_file": "benchmarks/benchmarks/benchmarks-agentic.yaml",
  "_cluster": "benchmarks",
  "attributes": {
    "displayName": "RE-Bench",
    "homepageUrl": "https://metr.org/AI_R_D_Evaluation_Report.pdf",
    "kind": "research-engineering",
    "targetsKind": "AgentVersion",
    "description": "METR's autonomous-research-engineering benchmark: time-bounded ML\nR&D tasks scored against expert human baselines to measure\nfrontier-agent ability to do self-directed research engineering.\n"
  },
  "outgoingEdges": [
    {
      "from": "benchmark:re-bench",
      "to": "skill-area:autonomous-research-engineering",
      "kind": "covers",
      "attributes": {
        "attributes": {
          "coverage": "full",
          "weight": 0.8
        }
      }
    },
    {
      "from": "benchmark:re-bench",
      "to": "skill-area:ml-fine-tuning",
      "kind": "covers",
      "attributes": {
        "attributes": {
          "coverage": "tangential",
          "weight": 0.2
        }
      }
    },
    {
      "from": "benchmark:re-bench",
      "to": "domain:machine-learning",
      "kind": "applies_to",
      "attributes": {}
    },
    {
      "from": "benchmark:re-bench",
      "to": "domain:scientific-discovery",
      "kind": "applies_to",
      "attributes": {}
    }
  ],
  "incomingEdges": []
}