{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YQG7SM6WLJYBL4BF6BPP4PVVF2","short_pith_number":"pith:YQG7SM6W","schema_version":"1.0","canonical_sha256":"c40df933d65a7015f025f05efe3eb52e850857ee5886314fc6151086727ad32b","source":{"kind":"arxiv","id":"2504.11713","version":3},"attestation_state":"computed","paper":{"title":"Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Aaron Havens, Anuroop Sriram, Benjamin Kurt Miller, Bing Yan, Bin Hu, Brandon Amos, Brandon Wood, Brian Karrer, Carles Domingo-Enrich, Daniel Levine, Guan-Horng Liu, Ricky T. Q. Chen, Xiang Fu","submitted_at":"2025-04-16T02:20:06Z","abstract_excerpt":"We introduce Adjoint Sampling, a highly scalable and efficient algorithm for learning diffusion processes that sample from unnormalized densities, or energy functions. It is the first on-policy approach that allows significantly more gradient updates than the number of energy evaluations and model samples, allowing us to scale to much larger problem settings than previously explored by similar methods. Our framework is theoretically grounded in stochastic optimal control and shares the same theoretical guarantees as Adjoint Matching, being able to train without the need for corrective measures"},"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":"2504.11713","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-04-16T02:20:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b76f1fd7fa2649fbff9cb2596e08ad5744bad98164f3d7f0ccea66f6b0ee27ee","abstract_canon_sha256":"f038049c7c9c86925c4f90b84d5e5f75154c74e964f5618c6c66df831d176960"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:22.192728Z","signature_b64":"VrlhIEE1H1gLdZGSbAGs/ofdxTU78ox/1zHxAR4dqNd2lvXJfUwbHHHZR6ve1OQPgP/jv7jalTOmUs1C9DaNBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c40df933d65a7015f025f05efe3eb52e850857ee5886314fc6151086727ad32b","last_reissued_at":"2026-07-05T11:11:22.192219Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:22.192219Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Aaron Havens, Anuroop Sriram, Benjamin Kurt Miller, Bing Yan, Bin Hu, Brandon Amos, Brandon Wood, Brian Karrer, Carles Domingo-Enrich, Daniel Levine, Guan-Horng Liu, Ricky T. Q. Chen, Xiang Fu","submitted_at":"2025-04-16T02:20:06Z","abstract_excerpt":"We introduce Adjoint Sampling, a highly scalable and efficient algorithm for learning diffusion processes that sample from unnormalized densities, or energy functions. It is the first on-policy approach that allows significantly more gradient updates than the number of energy evaluations and model samples, allowing us to scale to much larger problem settings than previously explored by similar methods. Our framework is theoretically grounded in stochastic optimal control and shares the same theoretical guarantees as Adjoint Matching, being able to train without the need for corrective measures"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.11713","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/2504.11713/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":"2504.11713","created_at":"2026-07-05T11:11:22.192279+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.11713v3","created_at":"2026-07-05T11:11:22.192279+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.11713","created_at":"2026-07-05T11:11:22.192279+00:00"},{"alias_kind":"pith_short_12","alias_value":"YQG7SM6WLJYB","created_at":"2026-07-05T11:11:22.192279+00:00"},{"alias_kind":"pith_short_16","alias_value":"YQG7SM6WLJYBL4BF","created_at":"2026-07-05T11:11:22.192279+00:00"},{"alias_kind":"pith_short_8","alias_value":"YQG7SM6W","created_at":"2026-07-05T11:11:22.192279+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22630","citing_title":"Scalable Maximum Entropy Reinforcement Learning for Diffusion Policies via Adjoint Matching","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10089","citing_title":"A Theory on Flow Matching with Neural Networks","ref_index":293,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25169","citing_title":"Laplace-Fisher Gate Identities for Optimal Matrix-Gated Blended Score Estimation","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18745","citing_title":"SURGE: Approximation and Training Free Particle Filter for Diffusion Surrogate","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25169","citing_title":"Laplace-Fisher Gate Identities for Optimal Matrix-Gated Blended Score Estimation","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23346","citing_title":"Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18745","citing_title":"SURGE: Approximation and Training Free Particle Filter for Diffusion Surrogate","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17850","citing_title":"Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measures","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2509.03726","citing_title":"Energy-Weighted Flow Matching: Unlocking Continuous Normalizing Flows for Efficient and Scalable Boltzmann Sampling","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2603.06082","citing_title":"Offline Materials Optimization with CliqueFlowmer","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21456","citing_title":"Tempered Sequential Monte Carlo for Trajectory and Policy Optimization with Differentiable Dynamics","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06583","citing_title":"Improved techniques for fine-tuning flow models via adjoint matching: a deterministic control pipeline","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21456","citing_title":"Tempered Sequential Monte Carlo for Trajectory and Policy Optimization with Differentiable Dynamics","ref_index":41,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2","json":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2.json","graph_json":"https://pith.science/api/pith-number/YQG7SM6WLJYBL4BF6BPP4PVVF2/graph.json","events_json":"https://pith.science/api/pith-number/YQG7SM6WLJYBL4BF6BPP4PVVF2/events.json","paper":"https://pith.science/paper/YQG7SM6W"},"agent_actions":{"view_html":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2","download_json":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2.json","view_paper":"https://pith.science/paper/YQG7SM6W","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.11713&json=true","fetch_graph":"https://pith.science/api/pith-number/YQG7SM6WLJYBL4BF6BPP4PVVF2/graph.json","fetch_events":"https://pith.science/api/pith-number/YQG7SM6WLJYBL4BF6BPP4PVVF2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2/action/storage_attestation","attest_author":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2/action/author_attestation","sign_citation":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2/action/citation_signature","submit_replication":"https://pith.science/pith/YQG7SM6WLJYBL4BF6BPP4PVVF2/action/replication_record"}},"created_at":"2026-07-05T11:11:22.192279+00:00","updated_at":"2026-07-05T11:11:22.192279+00:00"}