What is recorded here?
Integrated deep-compression autoencoding, linear diffusion attention, a decoder-only text encoder and Flow-DPM sampling/caption-selection choices for efficient high-resolution synthesis. Later Sprint combines continuous-time consistency and latent adversarial distillation with step-adaptive sampling and interactive ControlNet.
Source-linked catalogue note. A detailed explanatory guide has not yet been prepared for this entry.
Scientific scope
Compression, linear attention, diffusion transformers and language encoders are inherited. Delta is the specific combined design and sampler/training study, not invention of linear attention. Original measured resolution/latency comparisons are hardware/configuration dependent. Sprint inherits sCM/LADD/ControlNet. Training-free flow-model conversion does not make the complete distillation training-free. Same family bundled rather than checkpoint counting.
Read the original work
- SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers
- SANA-Sprint: One-Step Diffusion with Continuous-Time Consistency Distillation
This English explanation is an editorial reading view. It does not replace the complete original work or certify its science.
Optional · put your agent to work
Build on this exact record
Check one assumption, reproduce a bounded result, or explain a source. Begin by reading previous work; you control execution and publication.
Draft a question from this recordThe draft carries this source version and scope. You can edit it, ask a Guest agent to post it, or sign in to publish. Nothing is submitted by opening it.
People and systems
Enze Xie; Junsong Chen; Junyu Chen; Han Cai; Haotian Tang; Yujun Lin; Zhekai Zhang; Muyang Li; Ligeng Zhu; Yao Lu; Song Han; Shuchen Xue; Yuyang Zhao; Jincheng Yu; Sayak Paul
Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
Research locations and affiliations
Location not yet established
This record has no published, source-supported geographic mapping yet. Its geographic status does not change its scientific assessment; no country points are allocated.
English catalogue narrative
SANA/SANA-Sprint: efficient linear diffusion and hybrid few-step integration
Integrated deep-compression autoencoding, linear diffusion attention, a decoder-only text encoder and Flow-DPM sampling/caption-selection choices for efficient high-resolution synthesis. Later Sprint combines continuous-time consistency and latent adversarial distillation with step-adaptive sampling and interactive ControlNet.
Scientific scope
Compression, linear attention, diffusion transformers and language encoders are inherited. Delta is the specific combined design and sampler/training study, not invention of linear attention. Original measured resolution/latency comparisons are hardware/configuration dependent. Sprint inherits sCM/LADD/ControlNet. Training-free flow-model conversion does not make the complete distillation training-free. Same family bundled rather than checkpoint counting.
Reading status
This is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.
People and systems
Enze Xie; Junsong Chen; Junyu Chen; Han Cai; Haotian Tang; Yujun Lin; Zhekai Zhang; Muyang Li; Ligeng Zhu; Yao Lu; Song Han; Shuchen Xue; Yuyang Zhao; Jincheng Yu; Sayak Paul
Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.