{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OLNLCAE2LVYW7N3TFHDVH22XRJ","short_pith_number":"pith:OLNLCAE2","schema_version":"1.0","canonical_sha256":"72dab1009a5d716fb77329c753eb578a707fa36058097e0326e8415145acca4c","source":{"kind":"arxiv","id":"2506.08965","version":1},"attestation_state":"computed","paper":{"title":"GFRIEND: Generative Few-shot Reward Inference through EfficieNt DPO","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Huiyu Bai, Xuejiao Zhao, Yiyang Zhao","submitted_at":"2025-06-10T16:37:13Z","abstract_excerpt":"The ability to train high-performing reward models with few-shot data is critical for enhancing the efficiency and scalability of Reinforcement Learning from Human Feedback (RLHF). We propose a data augmentation and expansion framework that enables generative reward models trained on small datasets to achieve comparable performance to those trained on large-scale datasets. Traditional methods to train a generative reward model, such as Direct Preference Optimization (DPO), are constrained by inefficiencies in sample pairing and limited data diversity. This work introduces preference refinement"},"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":"2506.08965","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-06-10T16:37:13Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8e03303e6b8e991aee69e75af95d2ddcc0b34ec1e51f416d330362906e9e2b09","abstract_canon_sha256":"63b9c8b830f53aa6f707f6f551f20493ae1dc5cc5527e42076eaaf598f39d7c8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:16.142997Z","signature_b64":"mvq0ud9Wc3JyB53hLcNALMeB9CBXgVb2NM1cOtVxkPJlpDdynnM/zLj3iG3e3aZs642BQ7X9wJ4gwStvlDCrBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"72dab1009a5d716fb77329c753eb578a707fa36058097e0326e8415145acca4c","last_reissued_at":"2026-07-05T11:19:16.142513Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:16.142513Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GFRIEND: Generative Few-shot Reward Inference through EfficieNt DPO","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Huiyu Bai, Xuejiao Zhao, Yiyang Zhao","submitted_at":"2025-06-10T16:37:13Z","abstract_excerpt":"The ability to train high-performing reward models with few-shot data is critical for enhancing the efficiency and scalability of Reinforcement Learning from Human Feedback (RLHF). We propose a data augmentation and expansion framework that enables generative reward models trained on small datasets to achieve comparable performance to those trained on large-scale datasets. Traditional methods to train a generative reward model, such as Direct Preference Optimization (DPO), are constrained by inefficiencies in sample pairing and limited data diversity. This work introduces preference refinement"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.08965","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/2506.08965/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":"2506.08965","created_at":"2026-07-05T11:19:16.142576+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.08965v1","created_at":"2026-07-05T11:19:16.142576+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.08965","created_at":"2026-07-05T11:19:16.142576+00:00"},{"alias_kind":"pith_short_12","alias_value":"OLNLCAE2LVYW","created_at":"2026-07-05T11:19:16.142576+00:00"},{"alias_kind":"pith_short_16","alias_value":"OLNLCAE2LVYW7N3T","created_at":"2026-07-05T11:19:16.142576+00:00"},{"alias_kind":"pith_short_8","alias_value":"OLNLCAE2","created_at":"2026-07-05T11:19:16.142576+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/OLNLCAE2LVYW7N3TFHDVH22XRJ","json":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ.json","graph_json":"https://pith.science/api/pith-number/OLNLCAE2LVYW7N3TFHDVH22XRJ/graph.json","events_json":"https://pith.science/api/pith-number/OLNLCAE2LVYW7N3TFHDVH22XRJ/events.json","paper":"https://pith.science/paper/OLNLCAE2"},"agent_actions":{"view_html":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ","download_json":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ.json","view_paper":"https://pith.science/paper/OLNLCAE2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.08965&json=true","fetch_graph":"https://pith.science/api/pith-number/OLNLCAE2LVYW7N3TFHDVH22XRJ/graph.json","fetch_events":"https://pith.science/api/pith-number/OLNLCAE2LVYW7N3TFHDVH22XRJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ/action/storage_attestation","attest_author":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ/action/author_attestation","sign_citation":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ/action/citation_signature","submit_replication":"https://pith.science/pith/OLNLCAE2LVYW7N3TFHDVH22XRJ/action/replication_record"}},"created_at":"2026-07-05T11:19:16.142576+00:00","updated_at":"2026-07-05T11:19:16.142576+00:00"}