{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:IFFFKKEEWO523YN76RUL6YEBP4","short_pith_number":"pith:IFFFKKEE","schema_version":"1.0","canonical_sha256":"414a552884b3bbade1bff468bf60817f3272773873f6e20d4c138d9d99a1c31f","source":{"kind":"arxiv","id":"2211.01095","version":3},"attestation_state":"computed","paper":{"title":"DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"DPM-Solver++ generates high-quality guided samples from diffusion models in 15 to 20 steps","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Cheng Lu, Chongxuan Li, Fan Bao, Jianfei Chen, Jun Zhu, Yuhao Zhou","submitted_at":"2022-11-02T13:14:30Z","abstract_excerpt":"Diffusion probabilistic models (DPMs) have achieved impressive success in high-resolution image synthesis, especially in recent large-scale text-to-image generation applications. An essential technique for improving the sample quality of DPMs is guided sampling, which usually needs a large guidance scale to obtain the best sample quality. The commonly-used fast sampler for guided sampling is DDIM, a first-order diffusion ODE solver that generally needs 100 to 250 steps for high-quality samples. Although recent works propose dedicated high-order solvers and achieve a further speedup for samplin"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2211.01095","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-11-02T13:14:30Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"6fb94dc7b65c3a0888ddd207f6d8d4a4dc92e723d10cf97e60d056e20af0f2f3","abstract_canon_sha256":"2988b52dbb2d98e7850193f87431af48282a2efbb47e95cbc5a69c7fe45895b2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:48.597229Z","signature_b64":"Ru4000Yj7rQUyidLAfAJlcGJ5X4WpfW2ZCEkJIFTHSJ05M/seoRYl2duijt8kc+t3lnaZrfxT8bI+dJVvMNkBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"414a552884b3bbade1bff468bf60817f3272773873f6e20d4c138d9d99a1c31f","last_reissued_at":"2026-05-17T23:38:48.596639Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:48.596639Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"DPM-Solver++ generates high-quality guided samples from diffusion models in 15 to 20 steps","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Cheng Lu, Chongxuan Li, Fan Bao, Jianfei Chen, Jun Zhu, Yuhao Zhou","submitted_at":"2022-11-02T13:14:30Z","abstract_excerpt":"Diffusion probabilistic models (DPMs) have achieved impressive success in high-resolution image synthesis, especially in recent large-scale text-to-image generation applications. An essential technique for improving the sample quality of DPMs is guided sampling, which usually needs a large guidance scale to obtain the best sample quality. The commonly-used fast sampler for guided sampling is DDIM, a first-order diffusion ODE solver that generally needs 100 to 250 steps for high-quality samples. Although recent works propose dedicated high-order solvers and achieve a further speedup for samplin"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experiments show that DPM-Solver++ can generate high-quality samples within only 15 to 20 steps for guided sampling by pixel-space and latent-space DPMs.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the combination of data-prediction formulation and thresholding, plus the multistep variant, reliably eliminates the instability observed in prior high-order solvers when guidance scale is large.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"DPM-Solver++ enables high-quality guided sampling of diffusion models in 15-20 steps via data-prediction ODE solving and multistep stabilization.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"DPM-Solver++ generates high-quality guided samples from diffusion models in 15 to 20 steps","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c126f193e9d9447c50f649691bdb285e21a0491a25a0dbe8bf568caba049cfb7"},"source":{"id":"2211.01095","kind":"arxiv","version":3},"verdict":{"id":"9d033d3c-18b2-4189-adbc-48d57fd6ea7e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T07:50:42.794854Z","strongest_claim":"Experiments show that DPM-Solver++ can generate high-quality samples within only 15 to 20 steps for guided sampling by pixel-space and latent-space DPMs.","one_line_summary":"DPM-Solver++ enables high-quality guided sampling of diffusion models in 15-20 steps via data-prediction ODE solving and multistep stabilization.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the combination of data-prediction formulation and thresholding, plus the multistep variant, reliably eliminates the instability observed in prior high-order solvers when guidance scale is large.","pith_extraction_headline":"DPM-Solver++ generates high-quality guided samples from diffusion models in 15 to 20 steps"},"references":{"count":30,"sample":[{"doi":"","year":null,"title":"Estimating the optimal covariance with imperfect mean in diffusion probabilistic models","work_id":"d603de64-25fc-45d6-a2e3-facbb240bc7d","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"Classifier-free diffusion guidance","work_id":"fed77747-e4f2-4527-a3b8-a1822f33bbc2","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Gotta go fast when generating data with score-based models","work_id":"020cfe2a-ca4b-40ea-bc2d-53d06a2194b6","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"On fast sampling of diffusion probabilistic models","work_id":"28f4d037-fdd8-4254-92c7-3361c8b2633b","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Bilateral denoising diffusion models","work_id":"351c5468-87cd-4bbc-a9e6-f86d25051435","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":30,"snapshot_sha256":"50ce1fc1a9eb6b73630276fbccaa434463a744ca3472d82f0e7193c38a802199","internal_anchors":4},"formal_canon":{"evidence_count":2,"snapshot_sha256":"fab49c958ce80a0057f210f53de72bd2bc79983b390b015f135c2f6c79eeda0c"},"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":"2211.01095","created_at":"2026-05-17T23:38:48.596725+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.01095v3","created_at":"2026-05-17T23:38:48.596725+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.01095","created_at":"2026-05-17T23:38:48.596725+00:00"},{"alias_kind":"pith_short_12","alias_value":"IFFFKKEEWO52","created_at":"2026-05-18T12:33:33.725879+00:00"},{"alias_kind":"pith_short_16","alias_value":"IFFFKKEEWO523YN7","created_at":"2026-05-18T12:33:33.725879+00:00"},{"alias_kind":"pith_short_8","alias_value":"IFFFKKEE","created_at":"2026-05-18T12:33:33.725879+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":40,"internal_anchor_count":40,"sample":[{"citing_arxiv_id":"2304.09479","citing_title":"DiFaReli++: Diffusion Face Relighting with Consistent Cast Shadows","ref_index":38,"is_internal_anchor":true},{"citing_arxiv_id":"2311.04938","citing_title":"Improved DDIM Sampling with Moment Matching Gaussian Mixtures","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2405.18716","citing_title":"SketchDeco: Training-Free Latent Composition for Precise Sketch Colourisation","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2504.13109","citing_title":"UniEdit-Flow: Unleashing Inversion and Editing in the Era of Flow Models","ref_index":38,"is_internal_anchor":true},{"citing_arxiv_id":"2505.02242","citing_title":"Sampling-Aware Quantization for Diffusion Models","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2505.16024","citing_title":"Toward Theoretical Insights into Diffusion Trajectory Distillation via Operator Merging","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22586","citing_title":"A Tutorial on Diffusion Theory: From Differential Equations to Diffusion Models","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"2510.02307","citing_title":"NoiseShift: Resolution-Aware Noise Recalibration for Better Low-Resolution Image Generation","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2605.21042","citing_title":"Dynamic Video Generation: Shaping Video Generation Across Time and Space","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15684","citing_title":"ElasticDiT: Efficient Diffusion Transformers via Elastic Architecture and Sparse Attention for High-Resolution Image Generation on Mobile Devices","ref_index":57,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16126","citing_title":"Entropy Across the Bridge: Conditional-Marginal Discretization for Flow and Schr\\\"odinger Samplers","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17486","citing_title":"DyGRO-VLA: Cross-Task Scaling of Vision-Language-Action Models via Dynamic Grouped Residual Optimization","ref_index":92,"is_internal_anchor":true},{"citing_arxiv_id":"2406.03520","citing_title":"VideoPhy: Evaluating Physical Commonsense for Video Generation","ref_index":66,"is_internal_anchor":true},{"citing_arxiv_id":"2506.01666","citing_title":"Synthesis of discrete-continuous quantum circuits with multimodal diffusion models","ref_index":53,"is_internal_anchor":true},{"citing_arxiv_id":"2509.23727","citing_title":"AudioMoG: Guiding Audio Generation with Mixture-of-Guidance","ref_index":52,"is_internal_anchor":true},{"citing_arxiv_id":"2406.03736","citing_title":"Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Data","ref_index":71,"is_internal_anchor":true},{"citing_arxiv_id":"2211.15089","citing_title":"Continuous diffusion for categorical data","ref_index":57,"is_internal_anchor":true},{"citing_arxiv_id":"2511.20645","citing_title":"PixelDiT: Pixel Diffusion Transformers for Image Generation","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2512.13592","citing_title":"Image Diffusion Preview with Consistency Solver","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2601.06338","citing_title":"Circuit Mechanisms for Spatial Relation Generation in Diffusion Transformers","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2602.05449","citing_title":"DisCa: Accelerating Video Diffusion Transformers with Distillation-Compatible Learnable Feature Caching","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2410.10629","citing_title":"SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10983","citing_title":"TMPO: Trajectory Matching Policy Optimization for Diverse and Efficient Diffusion Alignment","ref_index":28,"is_internal_anchor":true},{"citing_arxiv_id":"2605.00836","citing_title":"From Euler to Dormand-Prince: ODE Solvers for Flow Matching Generative Models","ref_index":16,"is_internal_anchor":true},{"citing_arxiv_id":"2509.16117","citing_title":"DiffusionNFT: Online Diffusion Reinforcement with Forward Process","ref_index":16,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4","json":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4.json","graph_json":"https://pith.science/api/pith-number/IFFFKKEEWO523YN76RUL6YEBP4/graph.json","events_json":"https://pith.science/api/pith-number/IFFFKKEEWO523YN76RUL6YEBP4/events.json","paper":"https://pith.science/paper/IFFFKKEE"},"agent_actions":{"view_html":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4","download_json":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4.json","view_paper":"https://pith.science/paper/IFFFKKEE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.01095&json=true","fetch_graph":"https://pith.science/api/pith-number/IFFFKKEEWO523YN76RUL6YEBP4/graph.json","fetch_events":"https://pith.science/api/pith-number/IFFFKKEEWO523YN76RUL6YEBP4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4/action/storage_attestation","attest_author":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4/action/author_attestation","sign_citation":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4/action/citation_signature","submit_replication":"https://pith.science/pith/IFFFKKEEWO523YN76RUL6YEBP4/action/replication_record"}},"created_at":"2026-05-17T23:38:48.596725+00:00","updated_at":"2026-05-17T23:38:48.596725+00:00"}