{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:WAOZHUPQAPAK4XSKEVCXUTHGLV","short_pith_number":"pith:WAOZHUPQ","schema_version":"1.0","canonical_sha256":"b01d93d1f003c0ae5e4a25457a4ce65d4f6aceec2b43a9931f5827516673123f","source":{"kind":"arxiv","id":"2505.06079","version":1},"attestation_state":"computed","paper":{"title":"TREND: Tri-teaching for Robust Preference-based Reinforcement Learning with Demonstrations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.RO","authors_text":"Abhinav Shrivastava, Anubhav Gupta, Daniel Ekpo, Mara Levy, Ruijie Zheng, Shuaiyi Huang","submitted_at":"2025-05-09T14:22:43Z","abstract_excerpt":"Preference feedback collected by human or VLM annotators is often noisy, presenting a significant challenge for preference-based reinforcement learning that relies on accurate preference labels. To address this challenge, we propose TREND, a novel framework that integrates few-shot expert demonstrations with a tri-teaching strategy for effective noise mitigation. Our method trains three reward models simultaneously, where each model views its small-loss preference pairs as useful knowledge and teaches such useful pairs to its peer network for updating the parameters. Remarkably, our approach r"},"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":"2505.06079","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-05-09T14:22:43Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"c6562357fe48dbcdacf6967ac17b694e819b7d097387e3a8e9a0845dd0335203","abstract_canon_sha256":"0d259162feb71520dc6a58693725245aedc9de24ead405414e98a6a8f3eb68f7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:00:53.585172Z","signature_b64":"nDlZ+l2t2aVm40T4vqcA9PcjyhFXIZUXVcr9tdPIzrpv5Q0wy5182DG5PkuDju2JwwhAknLUOmyCiKvKaJyiDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b01d93d1f003c0ae5e4a25457a4ce65d4f6aceec2b43a9931f5827516673123f","last_reissued_at":"2026-07-05T11:00:53.584551Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:00:53.584551Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TREND: Tri-teaching for Robust Preference-based Reinforcement Learning with Demonstrations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.RO","authors_text":"Abhinav Shrivastava, Anubhav Gupta, Daniel Ekpo, Mara Levy, Ruijie Zheng, Shuaiyi Huang","submitted_at":"2025-05-09T14:22:43Z","abstract_excerpt":"Preference feedback collected by human or VLM annotators is often noisy, presenting a significant challenge for preference-based reinforcement learning that relies on accurate preference labels. To address this challenge, we propose TREND, a novel framework that integrates few-shot expert demonstrations with a tri-teaching strategy for effective noise mitigation. Our method trains three reward models simultaneously, where each model views its small-loss preference pairs as useful knowledge and teaches such useful pairs to its peer network for updating the parameters. Remarkably, our approach r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.06079","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/2505.06079/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":"2505.06079","created_at":"2026-07-05T11:00:53.584637+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.06079v1","created_at":"2026-07-05T11:00:53.584637+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.06079","created_at":"2026-07-05T11:00:53.584637+00:00"},{"alias_kind":"pith_short_12","alias_value":"WAOZHUPQAPAK","created_at":"2026-07-05T11:00:53.584637+00:00"},{"alias_kind":"pith_short_16","alias_value":"WAOZHUPQAPAK4XSK","created_at":"2026-07-05T11:00:53.584637+00:00"},{"alias_kind":"pith_short_8","alias_value":"WAOZHUPQ","created_at":"2026-07-05T11:00:53.584637+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07693","citing_title":"Selective Timestep Weighting and Advantage-Based Replay for Sample-Efficient Diffusion RLHF","ref_index":16,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV","json":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV.json","graph_json":"https://pith.science/api/pith-number/WAOZHUPQAPAK4XSKEVCXUTHGLV/graph.json","events_json":"https://pith.science/api/pith-number/WAOZHUPQAPAK4XSKEVCXUTHGLV/events.json","paper":"https://pith.science/paper/WAOZHUPQ"},"agent_actions":{"view_html":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV","download_json":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV.json","view_paper":"https://pith.science/paper/WAOZHUPQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.06079&json=true","fetch_graph":"https://pith.science/api/pith-number/WAOZHUPQAPAK4XSKEVCXUTHGLV/graph.json","fetch_events":"https://pith.science/api/pith-number/WAOZHUPQAPAK4XSKEVCXUTHGLV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV/action/storage_attestation","attest_author":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV/action/author_attestation","sign_citation":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV/action/citation_signature","submit_replication":"https://pith.science/pith/WAOZHUPQAPAK4XSKEVCXUTHGLV/action/replication_record"}},"created_at":"2026-07-05T11:00:53.584637+00:00","updated_at":"2026-07-05T11:00:53.584637+00:00"}