Pith. sign in

REVIEW 9 cited by

FlowAR: Scale-wise Autoregressive Image Generation Meets Flow Matching

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

arxiv 2412.15205 v1 pith:L62XB75G submitted 2024-12-19 cs.CV

FlowAR: Scale-wise Autoregressive Image Generation Meets Flow Matching

classification cs.CV
keywords scalenextpredictionflowarautoregressivedesigngenerationimage
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Autoregressive (AR) modeling has achieved remarkable success in natural language processing by enabling models to generate text with coherence and contextual understanding through next token prediction. Recently, in image generation, VAR proposes scale-wise autoregressive modeling, which extends the next token prediction to the next scale prediction, preserving the 2D structure of images. However, VAR encounters two primary challenges: (1) its complex and rigid scale design limits generalization in next scale prediction, and (2) the generator's dependence on a discrete tokenizer with the same complex scale structure restricts modularity and flexibility in updating the tokenizer. To address these limitations, we introduce FlowAR, a general next scale prediction method featuring a streamlined scale design, where each subsequent scale is simply double the previous one. This eliminates the need for VAR's intricate multi-scale residual tokenizer and enables the use of any off-the-shelf Variational AutoEncoder (VAE). Our simplified design enhances generalization in next scale prediction and facilitates the integration of Flow Matching for high-quality image synthesis. We validate the effectiveness of FlowAR on the challenging ImageNet-256 benchmark, demonstrating superior generation performance compared to previous methods. Codes will be available at \url{https://github.com/OliverRensu/FlowAR}.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 9 Pith papers

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

  1. Balancing Image Compression and Generation with Bootstrapped Tokenization

    cs.LG 2026-06 unverdicted novelty 7.0

    SelfBootTok decomposes image tokens into global and local groups via self-bootstrapped learning, enabling generators to use only global tokens for ~40% less computation and a new SOTA gFID of 1.56 with 64 tokens.

  2. Frequency-Forcing: From Scaling-as-Time to Soft Frequency Guidance

    cs.LG 2026-04 unverdicted novelty 7.0

    Frequency-Forcing guides pixel flow-matching with a data-derived low-frequency auxiliary stream to softly enforce scale-ordered generation, improving FID on ImageNet-256 over baselines.

  3. Drift-AR: Single-Step Visual Autoregressive Generation via Anti-Symmetric Drifting

    cs.CV 2026-03 unverdicted novelty 7.0

    Drift-AR achieves 3.8-5.5x speedup in AR-diffusion image models by using entropy to enable entropy-informed speculative decoding and single-step (1-NFE) anti-symmetric drifting decoding.

  4. Drift-AR: Single-Step Visual Autoregressive Generation via Anti-Symmetric Drifting

    cs.CV 2026-03 conditional novelty 6.5

    Per-position AR prediction entropy jointly drives speculative AR decoding and an anti-symmetric single-step drift decoder, yielding 3.8–5.5× faster hybrid visual generation without distillation.

  5. Beyond Token-Level Cross-Entropy: Fr\'echet Distributional Post-Training for Autoregressive Image Generation

    cs.CV 2026-08 conditional novelty 6.0

    FD-loss post-training with detached rollout replay and a probability-level straight-through estimator improves FID and FD_r6 across eight ImageNet configurations.

  6. Amortized Moment Matching for Visual Generation

    cs.LG 2026-07 accept novelty 6.0

    Amortized Fréchet Distance uses neural nets to match conditional means and covariances, yielding stronger one-step visual generators than explicit FD-loss or multi-step teachers.

  7. Edit the Bits, Diff the Codes: Bitwise Residual Editing for Visual Autoregressive Models

    cs.CV 2026-06 unverdicted novelty 6.0

    BitResEdit improves text alignment in VAR image editing by +1.07 CLIP on edited regions via bitwise source-negative guidance and gated residual code injection while preserving background.

  8. Trajectory Forcing: Structure-First Generation with Controllable Semantic Trajectories

    cs.CV 2026-06 unverdicted novelty 5.0

    Trajectory Forcing makes generative image synthesis trajectory-centric by organizing it into decodable semantic stages derived from clustered visual representations and trained with one-step flow-matching models.

  9. Rethinking Visual Autoregressive Sampling with Information-Grounding Guidance

    cs.CV 2025-09 conditional novelty 5.0

    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.