{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:RBGBX7E2TSLFPDAWTXRRVMF7IH","short_pith_number":"pith:RBGBX7E2","schema_version":"1.0","canonical_sha256":"884c1bfc9a9c96578c169de31ab0bf41da6fbb74ea2f5531924043dac31ffc2e","source":{"kind":"arxiv","id":"2408.17131","version":1},"attestation_state":"computed","paper":{"title":"VQ4DiT: Efficient Post-Training Vector Quantization for Diffusion Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Hong Gu, Juncan Deng, Kedong Xu, Kejie Huang, Shuaiting Li, Zeyu Wang","submitted_at":"2024-08-30T09:15:54Z","abstract_excerpt":"The Diffusion Transformers Models (DiTs) have transitioned the network architecture from traditional UNets to transformers, demonstrating exceptional capabilities in image generation. Although DiTs have been widely applied to high-definition video generation tasks, their large parameter size hinders inference on edge devices. Vector quantization (VQ) can decompose model weight into a codebook and assignments, allowing extreme weight quantization and significantly reducing memory usage. In this paper, we propose VQ4DiT, a fast post-training vector quantization method for DiTs. We found that tra"},"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":"2408.17131","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-08-30T09:15:54Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d5241e6d62a3ae8fa57904f5d298577df573fa513574c0f6fb0ee53f4f8ca9be","abstract_canon_sha256":"dedb9594766bc1d2bb6e88ba23d480a3b2ed225c5b57e45d81a69b2de32a5015"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:01:11.075215Z","signature_b64":"SQlQdDrxI30zQVt/VsqCB4bzXpHN6o0tKwJ8Y3ghQd2GKeOVCHr98rFkenzh8zkzbWIybNfCcfmq0jQvU1HeAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"884c1bfc9a9c96578c169de31ab0bf41da6fbb74ea2f5531924043dac31ffc2e","last_reissued_at":"2026-07-05T09:01:11.074822Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:01:11.074822Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VQ4DiT: Efficient Post-Training Vector Quantization for Diffusion Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Hong Gu, Juncan Deng, Kedong Xu, Kejie Huang, Shuaiting Li, Zeyu Wang","submitted_at":"2024-08-30T09:15:54Z","abstract_excerpt":"The Diffusion Transformers Models (DiTs) have transitioned the network architecture from traditional UNets to transformers, demonstrating exceptional capabilities in image generation. Although DiTs have been widely applied to high-definition video generation tasks, their large parameter size hinders inference on edge devices. Vector quantization (VQ) can decompose model weight into a codebook and assignments, allowing extreme weight quantization and significantly reducing memory usage. In this paper, we propose VQ4DiT, a fast post-training vector quantization method for DiTs. We found that tra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.17131","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/2408.17131/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":"2408.17131","created_at":"2026-07-05T09:01:11.074872+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.17131v1","created_at":"2026-07-05T09:01:11.074872+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.17131","created_at":"2026-07-05T09:01:11.074872+00:00"},{"alias_kind":"pith_short_12","alias_value":"RBGBX7E2TSLF","created_at":"2026-07-05T09:01:11.074872+00:00"},{"alias_kind":"pith_short_16","alias_value":"RBGBX7E2TSLFPDAW","created_at":"2026-07-05T09:01:11.074872+00:00"},{"alias_kind":"pith_short_8","alias_value":"RBGBX7E2","created_at":"2026-07-05T09:01:11.074872+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2411.05961","citing_title":"Aligned Vector Quantization for Edge-Cloud Collabrative Vision-Language Models","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH","json":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH.json","graph_json":"https://pith.science/api/pith-number/RBGBX7E2TSLFPDAWTXRRVMF7IH/graph.json","events_json":"https://pith.science/api/pith-number/RBGBX7E2TSLFPDAWTXRRVMF7IH/events.json","paper":"https://pith.science/paper/RBGBX7E2"},"agent_actions":{"view_html":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH","download_json":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH.json","view_paper":"https://pith.science/paper/RBGBX7E2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.17131&json=true","fetch_graph":"https://pith.science/api/pith-number/RBGBX7E2TSLFPDAWTXRRVMF7IH/graph.json","fetch_events":"https://pith.science/api/pith-number/RBGBX7E2TSLFPDAWTXRRVMF7IH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH/action/storage_attestation","attest_author":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH/action/author_attestation","sign_citation":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH/action/citation_signature","submit_replication":"https://pith.science/pith/RBGBX7E2TSLFPDAWTXRRVMF7IH/action/replication_record"}},"created_at":"2026-07-05T09:01:11.074872+00:00","updated_at":"2026-07-05T09:01:11.074872+00:00"}