{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XXKCHYWZVMTWAUFLFWDJIUOGY5","short_pith_number":"pith:XXKCHYWZ","schema_version":"1.0","canonical_sha256":"bdd423e2d9ab276050ab2d869451c6c7601f311dafd4cd488e1da2d169e1e2f3","source":{"kind":"arxiv","id":"2507.21504","version":1},"attestation_state":"computed","paper":{"title":"Evaluation and Benchmarking of LLM Agents: A Survey","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jane Lo, Mahmoud Mohammadi, Wendy Yip, Yipeng Li","submitted_at":"2025-07-29T04:57:02Z","abstract_excerpt":"The rise of LLM-based agents has opened new frontiers in AI applications, yet evaluating these agents remains a complex and underdeveloped area. This survey provides an in-depth overview of the emerging field of LLM agent evaluation, introducing a two-dimensional taxonomy that organizes existing work along (1) evaluation objectives -- what to evaluate, such as agent behavior, capabilities, reliability, and safety -- and (2) evaluation process -- how to evaluate, including interaction modes, datasets and benchmarks, metric computation methods, and tooling. In addition to taxonomy, we highlight "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2507.21504","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-29T04:57:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2d7031efddaae4f8955344570c7adf31dc2a92030ca4846352c79e868af4a6a2","abstract_canon_sha256":"627bae722f25417a48c4592ef04e985c60c3cbbedd85ef82a4d7038623bc0cc4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:45:01.347310Z","signature_b64":"CnwxY0gQwP+MAEwR5OZRMHGNV8/RkzkB3TnoQBlSODo+IUn0jAdci/LQ8XeWtDYfWt5qruthP6DtBbuSnFjwDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bdd423e2d9ab276050ab2d869451c6c7601f311dafd4cd488e1da2d169e1e2f3","last_reissued_at":"2026-07-05T11:45:01.346898Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:45:01.346898Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluation and Benchmarking of LLM Agents: A Survey","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jane Lo, Mahmoud Mohammadi, Wendy Yip, Yipeng Li","submitted_at":"2025-07-29T04:57:02Z","abstract_excerpt":"The rise of LLM-based agents has opened new frontiers in AI applications, yet evaluating these agents remains a complex and underdeveloped area. This survey provides an in-depth overview of the emerging field of LLM agent evaluation, introducing a two-dimensional taxonomy that organizes existing work along (1) evaluation objectives -- what to evaluate, such as agent behavior, capabilities, reliability, and safety -- and (2) evaluation process -- how to evaluate, including interaction modes, datasets and benchmarks, metric computation methods, and tooling. In addition to taxonomy, we highlight "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.21504","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2507.21504/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2507.21504","created_at":"2026-07-05T11:45:01.346953+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.21504v1","created_at":"2026-07-05T11:45:01.346953+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.21504","created_at":"2026-07-05T11:45:01.346953+00:00"},{"alias_kind":"pith_short_12","alias_value":"XXKCHYWZVMTW","created_at":"2026-07-05T11:45:01.346953+00:00"},{"alias_kind":"pith_short_16","alias_value":"XXKCHYWZVMTWAUFL","created_at":"2026-07-05T11:45:01.346953+00:00"},{"alias_kind":"pith_short_8","alias_value":"XXKCHYWZ","created_at":"2026-07-05T11:45:01.346953+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.11686","citing_title":"Layer-Isolated Evaluation: Gating the Deterministic Scaffold of a Production LLM Agent with a No-LLM, Regression-Locked Test Harness","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07805","citing_title":"Beyond Goodhart's Law: A Dynamic Benchmark for Evaluating Compliance in Multi-Agent Systems","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2603.20380","citing_title":"Herding CATs: ALARA for Agent Harness Engineering in Portable Composable Multi-Agent Teams","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16282","citing_title":"Taxonomy and Consistency Analysis of Safety Benchmarks for AI Agents","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16508","citing_title":"The Scaling Laws of Skills in LLM Agent Systems","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5","json":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5.json","graph_json":"https://pith.science/api/pith-number/XXKCHYWZVMTWAUFLFWDJIUOGY5/graph.json","events_json":"https://pith.science/api/pith-number/XXKCHYWZVMTWAUFLFWDJIUOGY5/events.json","paper":"https://pith.science/paper/XXKCHYWZ"},"agent_actions":{"view_html":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5","download_json":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5.json","view_paper":"https://pith.science/paper/XXKCHYWZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.21504&json=true","fetch_graph":"https://pith.science/api/pith-number/XXKCHYWZVMTWAUFLFWDJIUOGY5/graph.json","fetch_events":"https://pith.science/api/pith-number/XXKCHYWZVMTWAUFLFWDJIUOGY5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5/action/storage_attestation","attest_author":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5/action/author_attestation","sign_citation":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5/action/citation_signature","submit_replication":"https://pith.science/pith/XXKCHYWZVMTWAUFLFWDJIUOGY5/action/replication_record"}},"created_at":"2026-07-05T11:45:01.346953+00:00","updated_at":"2026-07-05T11:45:01.346953+00:00"}