{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2610.05532", "kind": "paper", "url": "https://arxiv.org/abs/2610.05532"}, {"id": "gh:pielord757/delegation-bench", "kind": "repo", "url": "https://github.com/PieLord757/delegation-bench"}], "authors": ["Shiva Pochampally"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2610.05532", "source_url": "https://arxiv.org/abs/2610.05532"}, "categories": ["General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "AI agents that send emails, edit files, and make purchases must decide when to act on their own and when to check with the user first."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2610.05532", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2610.05532", "title": "DelegationBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2610.05532", "modality": null, "name": "DelegationBench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_delegationbench_13990237", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "7dc184f05b7d05a20f25a7d9b94c7c296dcfdfce2e3ed1620ddb84929750af6e", "relation": "introduces", "review_state": "rule-auto-admitted", "reviewed_at": "2026-10-07T07:29:11.066253Z", "source_url": "https://arxiv.org/abs/2610.05532"}, "publisher": null, "released": "2026-10-04", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2610.05532", "source_url": "https://arxiv.org/abs/2610.05532"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2610-05532", "source": "claire_radar", "source_benchmark_id": "2610.05532", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "curation": {"decisionReason": "Automatically admitted from an explicit named benchmark release with evaluation-protocol evidence at or above the configured publication-confidence threshold.", "policy": "source-evidence-v2", "reviewedAt": "2026-10-07T07:29:11.066253Z", "state": "rule-auto-admitted"}, "dataStatus": "primary-source-indexed", "demo": false, "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce DelegationBench to test whether such scores can be trusted."}, "familyId": "bmf_5dee6762bec8", "firstSeenAt": "2026-10-06", "heat": null, "id": "bm_delegationbench_13990237", "indexedAt": "2026-10-07T07:28:43.628712Z", "industrySectors": [], "links": {"code": "https://github.com/PieLord757/delegation-bench", "data": null, "paper": "https://arxiv.org/abs/2610.05532", "pdf": "https://arxiv.org/pdf/2610.05532", "project": null, "report": "https://arxiv.org/abs/2610.05532"}, "metrics": [], "motivation": "AI agents that send emails, edit files, and make purchases must decide when to act on their own and when to check with the user first.", "name": "DelegationBench", "oneLine": "AI agents that send emails, edit files, and make purchases must decide when to act on their own and when to check with the user first.", "paperTitle": "DelegationBench: Measuring When AI Agents Should Ask Before Acting", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2610.05532", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-10-04", "source": {"authors": ["Shiva Pochampally"], "categories": ["cs.AI", "cs.CL", "cs.HC", "cs.LG"], "id": "2610.05532", "title": "DelegationBench: Measuring When AI Agents Should Ask Before Acting", "type": "arxiv", "url": "https://arxiv.org/abs/2610.05532"}, "sourceUpdatedAt": "2026-10-04T20:54:35Z", "topics": ["cs.AI"], "venueAttempts": []}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
