{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:FZLZBBCQ4TQ3LXU3TN5DY4RJMO","short_pith_number":"pith:FZLZBBCQ","schema_version":"1.0","canonical_sha256":"2e57908450e4e1b5de9b9b7a3c722963928e2b08a3def3ca960b4ce1375400b3","source":{"kind":"arxiv","id":"2311.13231","version":3},"attestation_state":"computed","paper":{"title":"Using Human Feedback to Fine-tune Diffusion Models without Any Reward Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Chunjiang Ge, Jiafei Lyu, Jian Tao, Jiaxin Chen, Kai Yang, Qimai Li, Weihan Shen, Xiaolong Zhu, Xiu Li","submitted_at":"2023-11-22T08:42:46Z","abstract_excerpt":"Using reinforcement learning with human feedback (RLHF) has shown significant promise in fine-tuning diffusion models. Previous methods start by training a reward model that aligns with human preferences, then leverage RL techniques to fine-tune the underlying models. However, crafting an efficient reward model demands extensive datasets, optimal architecture, and manual hyperparameter tuning, making the process both time and cost-intensive. The direct preference optimization (DPO) method, effective in fine-tuning large language models, eliminates the necessity for a reward model. However, the"},"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":"2311.13231","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-22T08:42:46Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"40aef761559b411b309bb39e2f24a0374f5a44b7c793db9bdc5e7a47efa41836","abstract_canon_sha256":"c89f4ffe584124e775b9f7bc6bb58d023bd9ca0961b5375b711e567aaa6241eb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:59:57.680280Z","signature_b64":"LVG2klJINvZjBAGXnynqZ6SecXiN6RWhCfSAvAwm2gNe3R/v3BU/RKziB5nyHuzhB5/WyNIjcZ1IxCtxAMEvBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2e57908450e4e1b5de9b9b7a3c722963928e2b08a3def3ca960b4ce1375400b3","last_reissued_at":"2026-07-05T07:59:57.679786Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:59:57.679786Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Using Human Feedback to Fine-tune Diffusion Models without Any Reward Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Chunjiang Ge, Jiafei Lyu, Jian Tao, Jiaxin Chen, Kai Yang, Qimai Li, Weihan Shen, Xiaolong Zhu, Xiu Li","submitted_at":"2023-11-22T08:42:46Z","abstract_excerpt":"Using reinforcement learning with human feedback (RLHF) has shown significant promise in fine-tuning diffusion models. Previous methods start by training a reward model that aligns with human preferences, then leverage RL techniques to fine-tune the underlying models. However, crafting an efficient reward model demands extensive datasets, optimal architecture, and manual hyperparameter tuning, making the process both time and cost-intensive. The direct preference optimization (DPO) method, effective in fine-tuning large language models, eliminates the necessity for a reward model. However, the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.13231","kind":"arxiv","version":3},"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/2311.13231/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":"2311.13231","created_at":"2026-07-05T07:59:57.679847+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.13231v3","created_at":"2026-07-05T07:59:57.679847+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.13231","created_at":"2026-07-05T07:59:57.679847+00:00"},{"alias_kind":"pith_short_12","alias_value":"FZLZBBCQ4TQ3","created_at":"2026-07-05T07:59:57.679847+00:00"},{"alias_kind":"pith_short_16","alias_value":"FZLZBBCQ4TQ3LXU3","created_at":"2026-07-05T07:59:57.679847+00:00"},{"alias_kind":"pith_short_8","alias_value":"FZLZBBCQ","created_at":"2026-07-05T07:59:57.679847+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"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":52,"is_internal_anchor":true},{"citing_arxiv_id":"2503.12575","citing_title":"BalancedDPO: Adaptive Multi-Metric Alignment","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17357","citing_title":"Dual-Diffusional Generative Fashion Recommendation","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06779","citing_title":"VASR: Variance-Aware Systematic Resampling for Reward-Guided Diffusion","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06779","citing_title":"VASR: Variance-Aware Systematic Resampling for Reward-Guided Diffusion","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO","json":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO.json","graph_json":"https://pith.science/api/pith-number/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/graph.json","events_json":"https://pith.science/api/pith-number/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/events.json","paper":"https://pith.science/paper/FZLZBBCQ"},"agent_actions":{"view_html":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO","download_json":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO.json","view_paper":"https://pith.science/paper/FZLZBBCQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.13231&json=true","fetch_graph":"https://pith.science/api/pith-number/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/graph.json","fetch_events":"https://pith.science/api/pith-number/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/action/storage_attestation","attest_author":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/action/author_attestation","sign_citation":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/action/citation_signature","submit_replication":"https://pith.science/pith/FZLZBBCQ4TQ3LXU3TN5DY4RJMO/action/replication_record"}},"created_at":"2026-07-05T07:59:57.679847+00:00","updated_at":"2026-07-05T07:59:57.679847+00:00"}