{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:4VZI4W62EECBYMC7VTQ76XJJ5S","short_pith_number":"pith:4VZI4W62","schema_version":"1.0","canonical_sha256":"e5728e5bda21041c305face1ff5d29eca7cc3d89a46cfd2886ddea772ec8ee2f","source":{"kind":"arxiv","id":"2508.02994","version":1},"attestation_state":"computed","paper":{"title":"When AIs Judge AIs: The Rise of Agent-as-a-Judge Evaluation for LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Fangyi Yu","submitted_at":"2025-08-05T01:42:25Z","abstract_excerpt":"As large language models (LLMs) grow in capability and autonomy, evaluating their outputs-especially in open-ended and complex tasks-has become a critical bottleneck. A new paradigm is emerging: using AI agents as the evaluators themselves. This \"agent-as-a-judge\" approach leverages the reasoning and perspective-taking abilities of LLMs to assess the quality and safety of other models, promising calable and nuanced alternatives to human evaluation. In this review, we define the agent-as-a-judge concept, trace its evolution from single-model judges to dynamic multi-agent debate frameworks, and "},"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":"2508.02994","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-08-05T01:42:25Z","cross_cats_sorted":[],"title_canon_sha256":"7b34d3e5a3acb4f551535ba257f02aa2ea3a0f97417db6875282900021efaa8f","abstract_canon_sha256":"73b64d0e578cf2c3be1c28493476f1b511f18065ae31b44d4c3cdf61323793ca"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:48:34.438548Z","signature_b64":"2+DcBfRPW5We2L6RD6klLKri5aWrZ+yxvCpgDDSmt+ZisqfaSnVOKP4N2Yv8y+cTYAzR/5XfnkgKLIUnVNS1AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e5728e5bda21041c305face1ff5d29eca7cc3d89a46cfd2886ddea772ec8ee2f","last_reissued_at":"2026-07-05T11:48:34.438053Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:48:34.438053Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When AIs Judge AIs: The Rise of Agent-as-a-Judge Evaluation for LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Fangyi Yu","submitted_at":"2025-08-05T01:42:25Z","abstract_excerpt":"As large language models (LLMs) grow in capability and autonomy, evaluating their outputs-especially in open-ended and complex tasks-has become a critical bottleneck. A new paradigm is emerging: using AI agents as the evaluators themselves. This \"agent-as-a-judge\" approach leverages the reasoning and perspective-taking abilities of LLMs to assess the quality and safety of other models, promising calable and nuanced alternatives to human evaluation. In this review, we define the agent-as-a-judge concept, trace its evolution from single-model judges to dynamic multi-agent debate frameworks, and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.02994","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/2508.02994/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":"2508.02994","created_at":"2026-07-05T11:48:34.438120+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.02994v1","created_at":"2026-07-05T11:48:34.438120+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.02994","created_at":"2026-07-05T11:48:34.438120+00:00"},{"alias_kind":"pith_short_12","alias_value":"4VZI4W62EECB","created_at":"2026-07-05T11:48:34.438120+00:00"},{"alias_kind":"pith_short_16","alias_value":"4VZI4W62EECBYMC7","created_at":"2026-07-05T11:48:34.438120+00:00"},{"alias_kind":"pith_short_8","alias_value":"4VZI4W62","created_at":"2026-07-05T11:48:34.438120+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20950","citing_title":"Power Systems Agent Benchmark: Executable Evaluation of AI Agents in Electric Power Engineering","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20950","citing_title":"Power Systems Agent Benchmark: Executable Evaluation of AI Agents in Electric Power Engineering","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18007","citing_title":"Semantic Reranking at Inference Time for Hard Examples in Rhetorical Role Labeling","ref_index":159,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S","json":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S.json","graph_json":"https://pith.science/api/pith-number/4VZI4W62EECBYMC7VTQ76XJJ5S/graph.json","events_json":"https://pith.science/api/pith-number/4VZI4W62EECBYMC7VTQ76XJJ5S/events.json","paper":"https://pith.science/paper/4VZI4W62"},"agent_actions":{"view_html":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S","download_json":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S.json","view_paper":"https://pith.science/paper/4VZI4W62","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.02994&json=true","fetch_graph":"https://pith.science/api/pith-number/4VZI4W62EECBYMC7VTQ76XJJ5S/graph.json","fetch_events":"https://pith.science/api/pith-number/4VZI4W62EECBYMC7VTQ76XJJ5S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S/action/storage_attestation","attest_author":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S/action/author_attestation","sign_citation":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S/action/citation_signature","submit_replication":"https://pith.science/pith/4VZI4W62EECBYMC7VTQ76XJJ5S/action/replication_record"}},"created_at":"2026-07-05T11:48:34.438120+00:00","updated_at":"2026-07-05T11:48:34.438120+00:00"}