{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:FNX3XV22LHCSYDZIQPKA47N6RZ","short_pith_number":"pith:FNX3XV22","schema_version":"1.0","canonical_sha256":"2b6fbbd75a59c52c0f2883d40e7dbe8e47d3bdcbcf4a9cfce014d654c9469540","source":{"kind":"arxiv","id":"2406.01125","version":1},"attestation_state":"computed","paper":{"title":"$\\Delta$-DiT: A Training-Free Acceleration Method Tailored for Diffusion Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chongjun Tu, Christos-Savvas Bouganis, Jianjian Cao, Mingzhu Shen, Pengtao Chen, Peng Ye, Tao Chen, Yiren Zhao","submitted_at":"2024-06-03T09:10:44Z","abstract_excerpt":"Diffusion models are widely recognized for generating high-quality and diverse images, but their poor real-time performance has led to numerous acceleration works, primarily focusing on UNet-based structures. With the more successful results achieved by diffusion transformers (DiT), there is still a lack of exploration regarding the impact of DiT structure on generation, as well as the absence of an acceleration framework tailored to the DiT architecture. To tackle these challenges, we conduct an investigation into the correlation between DiT blocks and image generation. Our findings reveal th"},"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":"2406.01125","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-06-03T09:10:44Z","cross_cats_sorted":[],"title_canon_sha256":"698adac612a24b4d610b7bfcbf818144429b4a8ded189386f7a5fb28d83664be","abstract_canon_sha256":"2826bd9094fc54fec22e6fd3391f17650c52c8688333bfd0219cefb8c8131b0b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:26:35.531901Z","signature_b64":"xFEpsERTAXtx6hk3Mp8Qxz82AIRCXyvkyOFpVJko2wE2Xejnv2Jz1mnXoaWLcfSQtOHVAaONV5DVX/aIPLrBDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2b6fbbd75a59c52c0f2883d40e7dbe8e47d3bdcbcf4a9cfce014d654c9469540","last_reissued_at":"2026-07-05T08:26:35.531431Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:26:35.531431Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"$\\Delta$-DiT: A Training-Free Acceleration Method Tailored for Diffusion Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chongjun Tu, Christos-Savvas Bouganis, Jianjian Cao, Mingzhu Shen, Pengtao Chen, Peng Ye, Tao Chen, Yiren Zhao","submitted_at":"2024-06-03T09:10:44Z","abstract_excerpt":"Diffusion models are widely recognized for generating high-quality and diverse images, but their poor real-time performance has led to numerous acceleration works, primarily focusing on UNet-based structures. With the more successful results achieved by diffusion transformers (DiT), there is still a lack of exploration regarding the impact of DiT structure on generation, as well as the absence of an acceleration framework tailored to the DiT architecture. To tackle these challenges, we conduct an investigation into the correlation between DiT blocks and image generation. Our findings reveal th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.01125","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/2406.01125/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":"2406.01125","created_at":"2026-07-05T08:26:35.531490+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.01125v1","created_at":"2026-07-05T08:26:35.531490+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.01125","created_at":"2026-07-05T08:26:35.531490+00:00"},{"alias_kind":"pith_short_12","alias_value":"FNX3XV22LHCS","created_at":"2026-07-05T08:26:35.531490+00:00"},{"alias_kind":"pith_short_16","alias_value":"FNX3XV22LHCSYDZI","created_at":"2026-07-05T08:26:35.531490+00:00"},{"alias_kind":"pith_short_8","alias_value":"FNX3XV22","created_at":"2026-07-05T08:26:35.531490+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":28,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26769","citing_title":"ResilPhase: Plug-and-Play Phase Mapping and Noise-Resilient Macro-Trajectory Extrapolation for Diffusion Acceleration","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20101","citing_title":"RFM-Editing 2: Text-Guided Audio Editing with Rectified Flow Matching and Coarse-to-Fine Diffusion Transformers","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20101","citing_title":"RFM-Editing 2: Text-Guided Audio Editing with Rectified Flow Matching and Coarse-to-Fine Diffusion Transformers","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06060","citing_title":"ReCache: Learning Budget-Aware Caching Schedules for Diffusion Models via REINFORCE","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31158","citing_title":"Light Interaction: Training-Free Inference Acceleration for Interactive Video World Models","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31026","citing_title":"OTCache: Optimal Transport for Geometry-Aware Caching in Diffusion Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29360","citing_title":"SAFE-DiT: Semantics-Aware Fast-path Execution for High-Resolution Diffusion Transformers","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26632","citing_title":"RT-Lynx: Putting the GEMM Sparsity In a Right Way for Diffusion Models","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27336","citing_title":"PARE: Pruning and Adaptive Routing for Efficient Video Generation","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23381","citing_title":"VDE: Training-Free Accelerating Rectified Flow Model via Velocity Decomposition and Estimation","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2405.14430","citing_title":"PipeFusion: Patch-level Pipeline Parallelism for Diffusion Transformers Inference","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22015","citing_title":"ORBIS: Output-Guided Token Reduction with Distribution-Aware Matching for Video Diffusion Acceleration","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18736","citing_title":"Spectral Progressive Diffusion for Efficient Image and Video Generation","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18736","citing_title":"Spectral Progressive Diffusion for Efficient Image and Video Generation","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2506.13456","citing_title":"Block-wise Adaptive Caching for Accelerating Diffusion Policy","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2602.05449","citing_title":"DisCa: Accelerating Video Diffusion Transformers with Distillation-Compatible Learnable Feature Caching","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2602.13357","citing_title":"AdaCorrection: Adaptive Offset Cache Correction for Accurate Diffusion Transformers","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01725","citing_title":"Motion-Aware Caching for Efficient Autoregressive Video Generation","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13316","citing_title":"Test-time Sparsity for Extreme Fast Action Diffusion","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02979","citing_title":"Not All Frames Deserve Full Computation: Accelerating Autoregressive Video Generation via Selective Computation and Predictive Extrapolation","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04018","citing_title":"1.x-Distill: Breaking the Diversity, Quality, and Efficiency Barrier in Distribution Matching Distillation","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11869","citing_title":"FIS-DiT: Breaking the Few-Step Video Inference Barrier via Training-Free Frame Interleaved Sparsity","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09003","citing_title":"FlashClear: Ultra-Fast Image Content Removal via Efficient Step Distillation and Feature Caching","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09003","citing_title":"FlashClear: Ultra-Fast Image Content Removal via Efficient Step Distillation and Feature Caching","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01725","citing_title":"Motion-Aware Caching for Efficient Autoregressive Video Generation","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ","json":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ.json","graph_json":"https://pith.science/api/pith-number/FNX3XV22LHCSYDZIQPKA47N6RZ/graph.json","events_json":"https://pith.science/api/pith-number/FNX3XV22LHCSYDZIQPKA47N6RZ/events.json","paper":"https://pith.science/paper/FNX3XV22"},"agent_actions":{"view_html":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ","download_json":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ.json","view_paper":"https://pith.science/paper/FNX3XV22","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.01125&json=true","fetch_graph":"https://pith.science/api/pith-number/FNX3XV22LHCSYDZIQPKA47N6RZ/graph.json","fetch_events":"https://pith.science/api/pith-number/FNX3XV22LHCSYDZIQPKA47N6RZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ/action/storage_attestation","attest_author":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ/action/author_attestation","sign_citation":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ/action/citation_signature","submit_replication":"https://pith.science/pith/FNX3XV22LHCSYDZIQPKA47N6RZ/action/replication_record"}},"created_at":"2026-07-05T08:26:35.531490+00:00","updated_at":"2026-07-05T08:26:35.531490+00:00"}