{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GTGE4LLL4NTVOYPFD6QT2BAMQI","short_pith_number":"pith:GTGE4LLL","schema_version":"1.0","canonical_sha256":"34cc4e2d6be3675761e51fa13d040c8233cb126ae2b5e2bfb7d803fd914d555c","source":{"kind":"arxiv","id":"2503.07565","version":7},"attestation_state":"computed","paper":{"title":"Inductive Moment Matching","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jiaming Song, Linqi Zhou, Stefano Ermon","submitted_at":"2025-03-10T17:37:39Z","abstract_excerpt":"Diffusion models and Flow Matching generate high-quality samples but are slow at inference, and distilling them into few-step models often leads to instability and extensive tuning. To resolve these trade-offs, we propose Inductive Moment Matching (IMM), a new class of generative models for one- or few-step sampling with a single-stage training procedure. Unlike distillation, IMM does not require pre-training initialization and optimization of two networks; and unlike Consistency Models, IMM guarantees distribution-level convergence and remains stable under various hyperparameters and standard"},"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":"2503.07565","kind":"arxiv","version":7},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-10T17:37:39Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"d5db7eb1e37935de0da88d2376da9452c42e3b8b9189383ed7ec878c7f169e40","abstract_canon_sha256":"b61c1b8790a23fb0649f993423ae0f320f7c630e755563832403596d150c81ae"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:03:22.040191Z","signature_b64":"gt0Dm8/7y6jfZjZA9eAxQ2wna84qWhfw2ri/aogdD++2GZ9zG3sfwBW9kkOdWQcnR7WsPJwDZD9HcCVxa9tDAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"34cc4e2d6be3675761e51fa13d040c8233cb126ae2b5e2bfb7d803fd914d555c","last_reissued_at":"2026-07-05T11:03:22.039732Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:03:22.039732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Inductive Moment Matching","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jiaming Song, Linqi Zhou, Stefano Ermon","submitted_at":"2025-03-10T17:37:39Z","abstract_excerpt":"Diffusion models and Flow Matching generate high-quality samples but are slow at inference, and distilling them into few-step models often leads to instability and extensive tuning. To resolve these trade-offs, we propose Inductive Moment Matching (IMM), a new class of generative models for one- or few-step sampling with a single-stage training procedure. Unlike distillation, IMM does not require pre-training initialization and optimization of two networks; and unlike Consistency Models, IMM guarantees distribution-level convergence and remains stable under various hyperparameters and standard"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.07565","kind":"arxiv","version":7},"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/2503.07565/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":"2503.07565","created_at":"2026-07-05T11:03:22.039790+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.07565v7","created_at":"2026-07-05T11:03:22.039790+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.07565","created_at":"2026-07-05T11:03:22.039790+00:00"},{"alias_kind":"pith_short_12","alias_value":"GTGE4LLL4NTV","created_at":"2026-07-05T11:03:22.039790+00:00"},{"alias_kind":"pith_short_16","alias_value":"GTGE4LLL4NTVOYPF","created_at":"2026-07-05T11:03:22.039790+00:00"},{"alias_kind":"pith_short_8","alias_value":"GTGE4LLL","created_at":"2026-07-05T11:03:22.039790+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.24888","citing_title":"DiffusionBench: On Holistic Evaluation of Diffusion Transformers","ref_index":139,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11138","citing_title":"First-Order Trajectory Matching: Fast Ensemble Predictions of Chaotic, Turbulent, Stochastic Systems","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11755","citing_title":"One-Step Generative Modeling via Wasserstein Gradient Flows","ref_index":75,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29920","citing_title":"Midpoint Generative Models","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17834","citing_title":"Stabilizing, Scaling & Enhancing MeanFlow for Large-scale Diffusion Distillation","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17546","citing_title":"Accelerating Redshift-Conditioned Galaxy Image Synthesis with One-step Generative Modeling","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15309","citing_title":"One Pass Is Not Enough: Recursive Latent Refinement for Generative Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2602.10764","citing_title":"Dual-End Consistency Model","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2603.01097","citing_title":"Understanding LoRA as Knowledge Memory: An Empirical Analysis","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11755","citing_title":"One-Step Generative Modeling via Wasserstein Gradient Flows","ref_index":71,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12183","citing_title":"DriftXpress: Faster Drifting Models via Projected RKHS Fields","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2505.13447","citing_title":"Mean Flows for One-step Generative Modeling","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06881","citing_title":"MENO: MeanFlow-Enhanced Neural Operators for Dynamical Systems","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06413","citing_title":"ODE-free Neural Flow Matching for One-Step Generative Modeling","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI","json":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI.json","graph_json":"https://pith.science/api/pith-number/GTGE4LLL4NTVOYPFD6QT2BAMQI/graph.json","events_json":"https://pith.science/api/pith-number/GTGE4LLL4NTVOYPFD6QT2BAMQI/events.json","paper":"https://pith.science/paper/GTGE4LLL"},"agent_actions":{"view_html":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI","download_json":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI.json","view_paper":"https://pith.science/paper/GTGE4LLL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.07565&json=true","fetch_graph":"https://pith.science/api/pith-number/GTGE4LLL4NTVOYPFD6QT2BAMQI/graph.json","fetch_events":"https://pith.science/api/pith-number/GTGE4LLL4NTVOYPFD6QT2BAMQI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI/action/storage_attestation","attest_author":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI/action/author_attestation","sign_citation":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI/action/citation_signature","submit_replication":"https://pith.science/pith/GTGE4LLL4NTVOYPFD6QT2BAMQI/action/replication_record"}},"created_at":"2026-07-05T11:03:22.039790+00:00","updated_at":"2026-07-05T11:03:22.039790+00:00"}