{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TKER44W6VNVB3LZ2C6T6C6YHTC","short_pith_number":"pith:TKER44W6","schema_version":"1.0","canonical_sha256":"9a891e72deab6a1daf3a17a7e17b0798b5e9cc3a1010d9b3578b3af4c7c9b37d","source":{"kind":"arxiv","id":"2508.14413","version":1},"attestation_state":"computed","paper":{"title":"Disentanglement in T-space for Faster and Distributed Training of Diffusion Models with Fewer Latent-states","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Aleix M. Martinez, Raghudeep Gadde, Rui Chen, Samarth Gupta","submitted_at":"2025-08-20T04:21:26Z","abstract_excerpt":"We challenge a fundamental assumption of diffusion models, namely, that a large number of latent-states or time-steps is required for training so that the reverse generative process is close to a Gaussian. We first show that with careful selection of a noise schedule, diffusion models trained over a small number of latent states (i.e. $T \\sim 32$) match the performance of models trained over a much large number of latent states ($T \\sim 1,000$). Second, we push this limit (on the minimum number of latent states required) to a single latent-state, which we refer to as complete disentanglement i"},"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":"2508.14413","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-20T04:21:26Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"60abfe24afa923b4d5a8253b02d5bf84281f6f3b56dffc6be78983669f90a1c5","abstract_canon_sha256":"c400c6b3d3fded7fdcf630f1739041ef9aaecc5ba2c633703bf271e75a57e059"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:56:31.447802Z","signature_b64":"WvxRXHQbEwjwcn91ctJbR7eDiSFZ2ZR4i0VonWx108VpUYGtPGHyXhj6W/np5oFXACKdQ8AEo7TWyUk/+YuECQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a891e72deab6a1daf3a17a7e17b0798b5e9cc3a1010d9b3578b3af4c7c9b37d","last_reissued_at":"2026-07-05T11:56:31.447421Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:56:31.447421Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Disentanglement in T-space for Faster and Distributed Training of Diffusion Models with Fewer Latent-states","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Aleix M. Martinez, Raghudeep Gadde, Rui Chen, Samarth Gupta","submitted_at":"2025-08-20T04:21:26Z","abstract_excerpt":"We challenge a fundamental assumption of diffusion models, namely, that a large number of latent-states or time-steps is required for training so that the reverse generative process is close to a Gaussian. We first show that with careful selection of a noise schedule, diffusion models trained over a small number of latent states (i.e. $T \\sim 32$) match the performance of models trained over a much large number of latent states ($T \\sim 1,000$). Second, we push this limit (on the minimum number of latent states required) to a single latent-state, which we refer to as complete disentanglement i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.14413","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/2508.14413/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":"2508.14413","created_at":"2026-07-05T11:56:31.447477+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.14413v1","created_at":"2026-07-05T11:56:31.447477+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.14413","created_at":"2026-07-05T11:56:31.447477+00:00"},{"alias_kind":"pith_short_12","alias_value":"TKER44W6VNVB","created_at":"2026-07-05T11:56:31.447477+00:00"},{"alias_kind":"pith_short_16","alias_value":"TKER44W6VNVB3LZ2","created_at":"2026-07-05T11:56:31.447477+00:00"},{"alias_kind":"pith_short_8","alias_value":"TKER44W6","created_at":"2026-07-05T11:56:31.447477+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/TKER44W6VNVB3LZ2C6T6C6YHTC","json":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC.json","graph_json":"https://pith.science/api/pith-number/TKER44W6VNVB3LZ2C6T6C6YHTC/graph.json","events_json":"https://pith.science/api/pith-number/TKER44W6VNVB3LZ2C6T6C6YHTC/events.json","paper":"https://pith.science/paper/TKER44W6"},"agent_actions":{"view_html":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC","download_json":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC.json","view_paper":"https://pith.science/paper/TKER44W6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.14413&json=true","fetch_graph":"https://pith.science/api/pith-number/TKER44W6VNVB3LZ2C6T6C6YHTC/graph.json","fetch_events":"https://pith.science/api/pith-number/TKER44W6VNVB3LZ2C6T6C6YHTC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC/action/storage_attestation","attest_author":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC/action/author_attestation","sign_citation":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC/action/citation_signature","submit_replication":"https://pith.science/pith/TKER44W6VNVB3LZ2C6T6C6YHTC/action/replication_record"}},"created_at":"2026-07-05T11:56:31.447477+00:00","updated_at":"2026-07-05T11:56:31.447477+00:00"}