{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FZDMR376QMS2W6P4XVIYZZ66BK","short_pith_number":"pith:FZDMR376","schema_version":"1.0","canonical_sha256":"2e46c8effe8325ab79fcbd518ce7de0abb5dc82b9673985e52dc154ab9874871","source":{"kind":"arxiv","id":"2607.14408","version":1},"attestation_state":"computed","paper":{"title":"Reward-Free Evolving Agents via Pairwise Validator","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiayun Wang, Minghao Liu, Wei Wei, Yu Wang","submitted_at":"2026-07-15T22:50:32Z","abstract_excerpt":"A self-evolving agentic loop repeatedly proposes a tweaked version of an agent (its prompt template or program) and accepts or rejects the change based on a per-iteration quality signal. Designing that signal is often the costly part of the project: a reliable scalar reward requires domain expertise and labeled examples that are themselves as expensive to assemble as the agent's underlying task. We propose replacing the scalar at the accept/reject gate with a pairwise validator: a frozen LLM that, given the parent and child candidate, returns a binary verdict on which is better. Pairwise judgm"},"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":"2607.14408","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-15T22:50:32Z","cross_cats_sorted":[],"title_canon_sha256":"2a48c3200728f0b4f5049d54d7ea3242f7a08fba720556f264111c12aa557cd0","abstract_canon_sha256":"30ece3f8961d5fdbade5569dfc4cfe54d63683dc5507f128ac8180ebe6d9ae79"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-17T00:21:11.068061Z","signature_b64":"KSSYEfbTspYmHzvytfyTn3apwOROGIIfJvuqL0X+vU3dmSzlopsBNz2c/Dkp3OVzAUN2Ik7GsyAh5/6UOQKoAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2e46c8effe8325ab79fcbd518ce7de0abb5dc82b9673985e52dc154ab9874871","last_reissued_at":"2026-07-17T00:21:11.067198Z","signature_status":"signed_v1","first_computed_at":"2026-07-17T00:21:11.067198Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reward-Free Evolving Agents via Pairwise Validator","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiayun Wang, Minghao Liu, Wei Wei, Yu Wang","submitted_at":"2026-07-15T22:50:32Z","abstract_excerpt":"A self-evolving agentic loop repeatedly proposes a tweaked version of an agent (its prompt template or program) and accepts or rejects the change based on a per-iteration quality signal. Designing that signal is often the costly part of the project: a reliable scalar reward requires domain expertise and labeled examples that are themselves as expensive to assemble as the agent's underlying task. We propose replacing the scalar at the accept/reject gate with a pairwise validator: a frozen LLM that, given the parent and child candidate, returns a binary verdict on which is better. Pairwise judgm"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.14408","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/2607.14408/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":"2607.14408","created_at":"2026-07-17T00:21:11.067653+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.14408v1","created_at":"2026-07-17T00:21:11.067653+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.14408","created_at":"2026-07-17T00:21:11.067653+00:00"},{"alias_kind":"pith_short_12","alias_value":"FZDMR376QMS2","created_at":"2026-07-17T00:21:11.067653+00:00"},{"alias_kind":"pith_short_16","alias_value":"FZDMR376QMS2W6P4","created_at":"2026-07-17T00:21:11.067653+00:00"},{"alias_kind":"pith_short_8","alias_value":"FZDMR376","created_at":"2026-07-17T00:21:11.067653+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/FZDMR376QMS2W6P4XVIYZZ66BK","json":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK.json","graph_json":"https://pith.science/api/pith-number/FZDMR376QMS2W6P4XVIYZZ66BK/graph.json","events_json":"https://pith.science/api/pith-number/FZDMR376QMS2W6P4XVIYZZ66BK/events.json","paper":"https://pith.science/paper/FZDMR376"},"agent_actions":{"view_html":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK","download_json":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK.json","view_paper":"https://pith.science/paper/FZDMR376","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.14408&json=true","fetch_graph":"https://pith.science/api/pith-number/FZDMR376QMS2W6P4XVIYZZ66BK/graph.json","fetch_events":"https://pith.science/api/pith-number/FZDMR376QMS2W6P4XVIYZZ66BK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK/action/storage_attestation","attest_author":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK/action/author_attestation","sign_citation":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK/action/citation_signature","submit_replication":"https://pith.science/pith/FZDMR376QMS2W6P4XVIYZZ66BK/action/replication_record"}},"created_at":"2026-07-17T00:21:11.067653+00:00","updated_at":"2026-07-17T00:21:11.067653+00:00"}