{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6IW4Z36OOPTX2HATIYXDKGTVRK","short_pith_number":"pith:6IW4Z36O","schema_version":"1.0","canonical_sha256":"f22dccefce73e77d1c13462e351a758a94ec4441179374bba62ee4ba1075d274","source":{"kind":"arxiv","id":"2504.15219","version":2},"attestation_state":"computed","paper":{"title":"EvalAgent: Discovering Implicit Evaluation Criteria from the Web","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chaitanya Malaviya, Greg Durrett, Junyi Jessy Li, Manya Wadhwa, Philippe Laban, Zayne Sprague","submitted_at":"2025-04-21T16:43:50Z","abstract_excerpt":"Evaluation of language model outputs on structured writing tasks is typically conducted with a number of desirable criteria presented to human evaluators or large language models (LLMs). For instance, on a prompt like \"Help me draft an academic talk on coffee intake vs research productivity\", a model response may be evaluated for criteria like accuracy and coherence. However, high-quality responses should do more than just satisfy basic task requirements. An effective response to this query should include quintessential features of an academic talk, such as a compelling opening, clear research"},"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":"2504.15219","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-21T16:43:50Z","cross_cats_sorted":[],"title_canon_sha256":"2a549495bfffd8bab9e1f3548da13cb2e80c34c82487520668a6de204c6b5671","abstract_canon_sha256":"bbc22af0e54967beedda2de96d1d029ac9bee744c97d75d2e4aabb0fb6bbad96"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:55:18.815461Z","signature_b64":"zs3zoWS2NGC2DJHqdKyMhXHbTg6NVk9+QhRGdAJQppOA+egDTb0tGOvYdw/oWvPprQDiMGeFUx3uQIjJLlcyBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f22dccefce73e77d1c13462e351a758a94ec4441179374bba62ee4ba1075d274","last_reissued_at":"2026-07-05T11:55:18.815029Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:55:18.815029Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"EvalAgent: Discovering Implicit Evaluation Criteria from the Web","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chaitanya Malaviya, Greg Durrett, Junyi Jessy Li, Manya Wadhwa, Philippe Laban, Zayne Sprague","submitted_at":"2025-04-21T16:43:50Z","abstract_excerpt":"Evaluation of language model outputs on structured writing tasks is typically conducted with a number of desirable criteria presented to human evaluators or large language models (LLMs). For instance, on a prompt like \"Help me draft an academic talk on coffee intake vs research productivity\", a model response may be evaluated for criteria like accuracy and coherence. However, high-quality responses should do more than just satisfy basic task requirements. An effective response to this query should include quintessential features of an academic talk, such as a compelling opening, clear research"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.15219","kind":"arxiv","version":2},"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/2504.15219/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":"2504.15219","created_at":"2026-07-05T11:55:18.815082+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.15219v2","created_at":"2026-07-05T11:55:18.815082+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.15219","created_at":"2026-07-05T11:55:18.815082+00:00"},{"alias_kind":"pith_short_12","alias_value":"6IW4Z36OOPTX","created_at":"2026-07-05T11:55:18.815082+00:00"},{"alias_kind":"pith_short_16","alias_value":"6IW4Z36OOPTX2HAT","created_at":"2026-07-05T11:55:18.815082+00:00"},{"alias_kind":"pith_short_8","alias_value":"6IW4Z36O","created_at":"2026-07-05T11:55:18.815082+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.21757","citing_title":"BehaviorSFT: Behavioral Token Conditioning for Clinical Agents Across the Proactivity Spectrum","ref_index":9,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK","json":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK.json","graph_json":"https://pith.science/api/pith-number/6IW4Z36OOPTX2HATIYXDKGTVRK/graph.json","events_json":"https://pith.science/api/pith-number/6IW4Z36OOPTX2HATIYXDKGTVRK/events.json","paper":"https://pith.science/paper/6IW4Z36O"},"agent_actions":{"view_html":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK","download_json":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK.json","view_paper":"https://pith.science/paper/6IW4Z36O","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.15219&json=true","fetch_graph":"https://pith.science/api/pith-number/6IW4Z36OOPTX2HATIYXDKGTVRK/graph.json","fetch_events":"https://pith.science/api/pith-number/6IW4Z36OOPTX2HATIYXDKGTVRK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK/action/storage_attestation","attest_author":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK/action/author_attestation","sign_citation":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK/action/citation_signature","submit_replication":"https://pith.science/pith/6IW4Z36OOPTX2HATIYXDKGTVRK/action/replication_record"}},"created_at":"2026-07-05T11:55:18.815082+00:00","updated_at":"2026-07-05T11:55:18.815082+00:00"}