REVIEW 7 cited by
Switti: Designing Scale-Wise Transformers for Text-to-Image Synthesis
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Switti: Designing Scale-Wise Transformers for Text-to-Image Synthesis
read the original abstract
This work presents Switti, a scale-wise transformer for text-to-image generation. We start by adapting an existing next-scale prediction autoregressive (AR) architecture to T2I generation, investigating and mitigating training stability issues in the process. Next, we argue that scale-wise transformers do not require causality and propose a non-causal counterpart facilitating ~21% faster sampling and lower memory usage while also achieving slightly better generation quality. Furthermore, we reveal that classifier-free guidance at high-resolution scales is often unnecessary and can even degrade performance. By disabling guidance at these scales, we achieve an additional sampling acceleration of ~32% and improve the generation of fine-grained details. Extensive human preference studies and automated evaluations show that Switti outperforms existing T2I AR models and competes with state-of-the-art T2I diffusion models while being up to 7x faster.
Forward citations
Cited by 7 Pith papers
-
Prompt-Guided Image Editing with Masked Logit Nudging in Visual Autoregressive Models
Masked Logit Nudging aligns visual autoregressive model logits with source token maps under target prompts inside cross-attention masks, delivering top image editing results on PIE benchmarks and strong reconstruction...
-
CGCE: Classifier-Guided Concept Erasure in Generative Models
A classifier on text embeddings detects unsafe prompts and uses gradient descent to steer the embedding to a safe region, achieving state-of-the-art concept erasure without changing the generative model's weights.
-
Revisiting Autoregressive Models for Generative Image Classification
Order-marginalized any-order AR models (RandAR) outperform diffusion generative classifiers on ImageNet and OOD sets and match strong SSL models at far lower cost.
-
UniGen-AR: Unifying Visual Generation with Auto-Regressive Modeling
An MLLM-conditioned next-scale VAR decoder handles 15+ unified visual generation tasks with competitive quality and substantially lower latency than diffusion baselines.
-
Uni-AdaVD: Universal Concept Erasure for Visual Generation via Orthogonal Value Decomposition
A training-free method erases concepts from diverse visual generators by orthogonally projecting attention value vectors away from encoder-aware target representations, with layer-adaptive erasure strength.
-
Knowledge Distillation for Visual Autoregressive Models
VarKD is a distillation framework for visual AR models that uses student samples and selective teacher supervision to reduce token ambiguity, outperforming prior baselines on ImageNet.
-
Rethinking Visual Autoregressive Sampling with Information-Grounding Guidance
IGG, an attention-based reweighting of classifier-free guidance, concentrates guidance on important tokens and modestly improves FID/IS in scale-wise autoregressive image generation.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.