Pith. sign in

REVIEW 9 cited by

HART: Efficient Visual Generation with Hybrid Autoregressive Transformer

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

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

arxiv 2410.10812 v1 pith:FC5OAH3F submitted 2024-10-14 cs.CV cs.AIcs.LG

Haotian Tang , Yecheng Wu , Shang Yang , Enze Xie , Junsong Chen , Junyu Chen , Zhuoyang Zhang , Han Cai
show 2 more authors
Yao Lu Song Han
This is my paper
classification cs.CVcs.AIcs.LG
keywords discretegenerationharthybridautoregressivecontinuousdiffusionmodels
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
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.

Discussion (0). Sign in to comment.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Detail Continuation over a Trustworthy Coarse Scale for Autoregressive Super-Resolution

    cs.CV 2026-08 conditional novelty 6.0 of 10

    In VAR-based super-resolution, K2N predicts the first three coarse scales in parallel from the low-resolution input and generates only the remaining fine scales autoregressively, reducing hallucination while staying c...

  2. Na\"ive PAINE: Lightweight Text-to-Image Generation Improvement with Prompt Evaluation

    cs.CV 2026-03 conditional novelty 6.0 of 10

    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.

  3. Discrete Noise Inversion for Next-scale Autoregressive Text-based Image Editing

    cs.CV 2025-09 conditional novelty 6.0 of 10

    VARIN uses a Location-aware Argmax Inversion pseudo-inverse of Gumbel-max sampling to extract editable discrete noises, enabling training-free prompt-guided editing for visual autoregressive models.

  4. Rethinking Discrete Tokens: Treating Them as Conditions for Continuous Autoregressive Image Synthesis

    cs.CV 2025-07 conditional novelty 6.0 of 10

    DisCon treats discrete image tokens as conditioning signals rather than targets, letting a continuous autoregressive model refine details and reach gFID 1.38 on ImageNet-256.

  5. Argus-Unified: Towards A Compact and Economical Unified Model for Image Understanding and Generation

    cs.CV 2026-07 conditional novelty 5.0 of 10

    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.

  6. Adaptive Visual Autoregressive Acceleration via Dual-Linkage Entropy Analysis

    cs.CV 2026-02 conditional novelty 5.0 of 10

    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...

  7. DC-AR: Efficient Masked Autoregressive Image Generation with Deep Compression Hybrid Tokenizer

    cs.CV 2025-07 conditional novelty 5.0 of 10

    DC-AR generates 512x512 images in 12 masked autoregressive steps plus 20 diffusion refinement steps, using a 32x compressed 2D tokenizer, and reports gFID 5.49 on MJHQ-30K.

  8. CycleVAR: Repurposing Autoregressive Model for Unsupervised One-Step Image Translation

    cs.CV 2025-06 conditional novelty 5.0 of 10

    CycleVAR adapts a pretrained visual autoregressive model to unpaired image translation using softmax-relaxed quantization and source-token prefixes, achieving FID scores competitive with CycleGAN-Turbo.

  9. FlowSonic: Stable Zero-Shot Music Editing via High-Order Trajectory Integration

    cs.SD 2026-07 reject novelty 4.0 of 10

    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.

Pith tools