REVIEW 16 cited by
HART: Efficient Visual Generation with Hybrid Autoregressive Transformer
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
HART: Efficient Visual Generation with Hybrid Autoregressive Transformer
read the original abstract
We introduce Hybrid Autoregressive Transformer (HART), an autoregressive (AR) visual generation model capable of directly generating 1024x1024 images, rivaling diffusion models in image generation quality. Existing AR models face limitations due to the poor image reconstruction quality of their discrete tokenizers and the prohibitive training costs associated with generating 1024px images. To address these challenges, we present the hybrid tokenizer, which decomposes the continuous latents from the autoencoder into two components: discrete tokens representing the big picture and continuous tokens representing the residual components that cannot be represented by the discrete tokens. The discrete component is modeled by a scalable-resolution discrete AR model, while the continuous component is learned with a lightweight residual diffusion module with only 37M parameters. Compared with the discrete-only VAR tokenizer, our hybrid approach improves reconstruction FID from 2.11 to 0.30 on MJHQ-30K, leading to a 31% generation FID improvement from 7.85 to 5.38. HART also outperforms state-of-the-art diffusion models in both FID and CLIP score, with 4.5-7.7x higher throughput and 6.9-13.4x lower MACs. Our code is open sourced at https://github.com/mit-han-lab/hart.
Forward citations
Cited by 16 Pith papers
-
DeVAR: Low-Dose CT Denoising via Visual Autoregressive Modeling
DeVAR is the first application of visual autoregressive modeling to low-dose CT denoising, using next-scale token prediction, a residual refiner, and hybrid discrete-continuous decoding to outperform prior methods on ...
-
HACK++: Towards More Effective Head-Aware Key-Value Compression for Efficient Visual Autoregressive Modeling
HACK++ is a head-aware KV cache compression framework for VAR models that decouples current-scale attention from historical cache under adaptive per-head budgets to achieve near-lossless generation at 30% attention an...
-
Where to Refine, When to Stop: Rethinking Redundancy via Latent Discrepancy for Efficient Visual Autoregressive Generation
LD-Pruning applies latent discrepancy to prune tokens and adaptively skip unconditional branches in VAR models for up to 2.35x faster inference with preserved quality.
-
ExtraVAR: Stage-Aware RoPE Remapping for Resolution Extrapolation in Visual Autoregressive Models
ExtraVAR enables resolution extrapolation in visual autoregressive models by stage-aware RoPE remapping and entropy-driven attention scaling, suppressing repetition and detail loss.
-
Autoregressive Visual Generation Needs a Prologue
Prologue introduces dedicated prologue tokens to decouple generation and reconstruction in AR visual models, significantly improving generation FID scores on ImageNet while maintaining reconstruction quality.
-
Autoregressive Visual Generation Needs a Prologue
Prologue adds a small set of learnable tokens trained exclusively with AR cross-entropy loss to decouple generation from reconstruction in autoregressive visual models, yielding lower gFID on ImageNet 256x256.
-
Discrete Stochastic Localization for Non-autoregressive Generation
Discrete Stochastic Localization lets a single trained network support an entire family of per-token SNR paths for discrete sequence generation, with masked diffusion as a special case, and improves MAUVE scores when ...
-
PacTure: Efficient PBR Texture Generation on Packed Views with Visual Autoregressive Models
PacTure uses view packing and next-scale autoregressive prediction to generate consistent multi-view PBR textures faster than prior sequential or cross-attention methods.
-
Concept Removal for Frontier Image Generative Models
A transcoder-based in-place replacement of the bottleneck layer enables selective concept removal in modern diffusion and autoregressive image models without degrading output quality.
-
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.
-
FlashAR: Efficient Post-Training Acceleration for Autoregressive Image Generation
FlashAR accelerates autoregressive image generation up to 22.9x by post-training a pre-trained raster-scan model with a complementary vertical head and dynamic fusion for two-way next-token prediction.
-
FlashAR: Efficient Post-Training Acceleration for Autoregressive Image Generation
FlashAR achieves up to 22.9x speedup in 512x512 autoregressive image generation by post-training a pre-trained model with a complementary vertical head and dynamic fusion using only 0.05% of original training data.
-
Na\"ive PAINE: Lightweight Text-to-Image Generation Improvement with Prompt Evaluation
A lightweight predictor ranks initial noises by expected human-preference score for a prompt, selecting the best few for diffusion generation and reporting prompt difficulty.
-
Argus-Unified: Towards A Compact and Economical Unified Model for Image Understanding and Generation
A compact unified model that reuses a frozen VLM encoder and hybrid continuous/discrete tokens reaches competitive image understanding and generation with 15.6M training images and about $2,000 in compute.
-
Adaptive Visual Autoregressive Acceleration via Dual-Linkage Entropy Analysis
A training-free entropy-guided token-pruning framework accelerates VAR image generation up to 2.9× with negligible benchmark loss by activating pruning at an adaptive entropy-growth inflection point and adjusting rati...
-
FlowSonic: Stable Zero-Shot Music Editing via High-Order Trajectory Integration
FlowSonic combines deterministic rectified-flow inversion, cached cross-attention injection, and a 'seeded' third-order Adams-Bashforth solver to report better timbre and genre edits on small datasets.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.