reports/mind2web-v004.json

{
  "source": "osunlp/Mind2Web",
  "revision": "6314166657eec4aa0e22c00f8d801e609ce8e80f",
  "file": "train_10.json",
  "source_sha256": "182542d7947b3fa9e90fc57a3d82d4d8f2997ca5a06664217720d7a78a956e33",
  "license": "CC-BY-4.0",
  "attribution": "Deng et al., Mind2Web: Towards a Generalist Agent for the Web, 2023, arXiv:2306.06070",
  "checkpoint": "models/v004-contextual",
  "tasks": 9,
  "websites": 3,
  "total_steps": 49,
  "scorable_steps": 46,
  "unscorable_steps": 3,
  "operations": {
    "C": 42,
    "T": 3,
    "O": 1
  },
  "candidate_recall": 0.47826087474823,
  "action_accuracy": 0.021739130839705467,
  "target_accuracy": 0.021739130839705467,
  "joint_accuracy": 0.021739130839705467,
  "browser_execution": false,
  "training_on_source": false,
  "limitations": "Small training-shard diagnostic, NOT official held-out benchmark. Approximate HTML names/roles; no history; 24-token goal truncation. No raw page or goal text persisted in report."
}