{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7E5A6R7EFGGLXXU5D6PPZ657MR","short_pith_number":"pith:7E5A6R7E","schema_version":"1.0","canonical_sha256":"f93a0f47e4298cbbde9d1f9efcfbbf644142f5ed9a3c5372bfd7cc8845cd551b","source":{"kind":"arxiv","id":"2608.02444","version":1},"attestation_state":"computed","paper":{"title":"ParEvalLayer: When Partial LLM-Agent Evaluations Support a Decision","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bonan Shen, Wei-Jung Huang","submitted_at":"2026-08-03T16:22:51Z","abstract_excerpt":"LLM-agent evaluations often produce task outcomes long before the full benchmark run is complete. A partial score is tempting to report, but it does not show whether the observed tasks support the same conclusion as the completed evaluation. Early tasks can omit important parts of a benchmark, running cheaper tasks first can distort the observed sample, and a rule that decides only easy pairs can appear accurate while leaving many comparisons unresolved. We introduce ParEvalLayer, a decision layer that reads paired outcomes for two agent systems and a comparison policy chosen in advance. For e"},"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":"2608.02444","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-03T16:22:51Z","cross_cats_sorted":[],"title_canon_sha256":"445852220788ffbe4f79c13e7180762961df2afd7e0744f4ad58a17803b743c2","abstract_canon_sha256":"e256002983e84b336791fd9fd395415325658384adbb2120a435ff03b576b8f2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:43:31.225505Z","signature_b64":"kazDKQDLEtcEWU51ZPIZiawupcYzy6IUqB8HnySt9bKzZMTvD3e0NraUaTLxd9DU5f2d3ynslg3d5a30H9QACw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f93a0f47e4298cbbde9d1f9efcfbbf644142f5ed9a3c5372bfd7cc8845cd551b","last_reissued_at":"2026-08-04T02:43:31.223074Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:43:31.223074Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ParEvalLayer: When Partial LLM-Agent Evaluations Support a Decision","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bonan Shen, Wei-Jung Huang","submitted_at":"2026-08-03T16:22:51Z","abstract_excerpt":"LLM-agent evaluations often produce task outcomes long before the full benchmark run is complete. A partial score is tempting to report, but it does not show whether the observed tasks support the same conclusion as the completed evaluation. Early tasks can omit important parts of a benchmark, running cheaper tasks first can distort the observed sample, and a rule that decides only easy pairs can appear accurate while leaving many comparisons unresolved. We introduce ParEvalLayer, a decision layer that reads paired outcomes for two agent systems and a comparison policy chosen in advance. For e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.02444","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/2608.02444/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":"2608.02444","created_at":"2026-08-04T02:43:31.225402+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.02444v1","created_at":"2026-08-04T02:43:31.225402+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.02444","created_at":"2026-08-04T02:43:31.225402+00:00"},{"alias_kind":"pith_short_12","alias_value":"7E5A6R7EFGGL","created_at":"2026-08-04T02:43:31.225402+00:00"},{"alias_kind":"pith_short_16","alias_value":"7E5A6R7EFGGLXXU5","created_at":"2026-08-04T02:43:31.225402+00:00"},{"alias_kind":"pith_short_8","alias_value":"7E5A6R7E","created_at":"2026-08-04T02:43:31.225402+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR","json":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR.json","graph_json":"https://pith.science/api/pith-number/7E5A6R7EFGGLXXU5D6PPZ657MR/graph.json","events_json":"https://pith.science/api/pith-number/7E5A6R7EFGGLXXU5D6PPZ657MR/events.json","paper":"https://pith.science/paper/7E5A6R7E"},"agent_actions":{"view_html":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR","download_json":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR.json","view_paper":"https://pith.science/paper/7E5A6R7E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.02444&json=true","fetch_graph":"https://pith.science/api/pith-number/7E5A6R7EFGGLXXU5D6PPZ657MR/graph.json","fetch_events":"https://pith.science/api/pith-number/7E5A6R7EFGGLXXU5D6PPZ657MR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR/action/storage_attestation","attest_author":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR/action/author_attestation","sign_citation":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR/action/citation_signature","submit_replication":"https://pith.science/pith/7E5A6R7EFGGLXXU5D6PPZ657MR/action/replication_record"}},"created_at":"2026-08-04T02:43:31.225402+00:00","updated_at":"2026-08-04T02:43:31.225402+00:00"}