Pith. sign in

REVIEW 11 cited by

Scalable Adaptive Computation for Iterative Generation

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 2212.11972 v2 pith:2DZKR47K submitted 2022-12-22 cs.LG cs.CVcs.NE

Allan Jabri , David Fleet , Ting Chen This is my paper
classification cs.LGcs.CVcs.NE
keywords computationdatalatentgenerationdiffusionrinsroutingtokens
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Natural data is redundant yet predominant architectures tile computation uniformly across their input and output space. We propose the Recurrent Interface Networks (RINs), an attention-based architecture that decouples its core computation from the dimensionality of the data, enabling adaptive computation for more scalable generation of high-dimensional data. RINs focus the bulk of computation (i.e. global self-attention) on a set of latent tokens, using cross-attention to read and write (i.e. route) information between latent and data tokens. Stacking RIN blocks allows bottom-up (data to latent) and top-down (latent to data) feedback, leading to deeper and more expressive routing. While this routing introduces challenges, this is less problematic in recurrent computation settings where the task (and routing problem) changes gradually, such as iterative generation with diffusion models. We show how to leverage recurrence by conditioning the latent tokens at each forward pass of the reverse diffusion process with those from prior computation, i.e. latent self-conditioning. RINs yield state-of-the-art pixel diffusion models for image and video generation, scaling to 1024X1024 images without cascades or guidance, while being domain-agnostic and up to 10X more efficient than 2D and 3D U-Nets.

Discussion (0). Sign in to comment.

Forward citations

Cited by 11 Pith papers

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

  1. SPAE: Spectrally Guided Autoencoder for Pretrained Visual Latents

    cs.CV 2026-08 conditional novelty 7.0 of 10

    SPAE compresses pretrained vision latents into a 64-channel bottleneck with channel-wise masking, improving image generation while keeping reconstruction and semantic understanding strong.

  2. A Frozen Pixel-Space Diffusion Model Can Guide Itself with Its Own Samples

    cs.CV 2026-07 conditional novelty 7.0 of 10

    Training a small adapter on a frozen pixel diffusion model's own samples and extrapolating the final prediction away from the adapter's intermediate prediction improves FID on ImageNet.

  3. WaiT for the Signal: Simple Frequency-Aware Flow-Matching

    cs.CV 2026-07 conditional novelty 6.0 of 10

    WaiT delays high-frequency wavelet bands in flow-matching image generation until coarse structure emerges, improving quality and cutting compute, with a reported SOTA FID of 1.30 on ImageNet 512.

  4. MIMFlow: Integrating Masked Image Modeling with Normalizing Flows for End-to-End Image Generation

    cs.CV 2026-06 unverdicted novelty 6.0 of 10

    MIMFlow uses a VAE on masked images to feed semantic latents to a normalizing flow while a decoder handles high-frequency details, reporting FID 2.50 and 71.3% linear probing on ImageNet 256x256 with 128 tokens.

  5. SRC-Flow: Compact Semantic Representations Enable Normalizing Flows for Image Generation

    cs.CV 2026-05 unverdicted novelty 6.0 of 10

    SRC-Flow compresses RAE features via a Semantic Representation Compressor into a low-dimensional space, enabling normalizing flows to reach gFID 1.65 on ImageNet 256x256 and 2.07 on 512x512 while retaining exact likelihoods.

  6. ELT: Elastic Looped Transformers for Visual Generation

    cs.CV 2026-04 unverdicted novelty 6.0 of 10

    Weight-shared looped transformers trained with intra-loop self-distillation match MaskGIT-class FID/FVD at roughly 4x fewer parameters and support any-time inference across loop counts.

  7. MetaState: Persistent Working Memory Enhances Reasoning in Discrete Diffusion Language Models

    cs.CL 2026-03 conditional novelty 6.0 of 10

    Adding a 0.6%-parameter recurrent working memory to frozen discrete diffusion LLMs yields an average 4.5-percentage-point gain on GSM8K, MATH-500, HumanEval, and MBPP.

  8. PixNerd: Pixel Neural Field Diffusion

    cs.CV 2025-07 conditional novelty 6.0 of 10

    PixNerd is a single-stage pixel-space diffusion transformer that uses predicted neural field weights to decode large patches, reaching 2.15 FID on ImageNet 256 without a VAE.

  9. DMQ: Dissecting Outliers of Diffusion Models for Post-Training Quantization

    cs.CV 2025-07 conditional novelty 6.0 of 10

    A post-training quantization method that combines learned channel scaling and power-of-two scaling keeps diffusion image quality high at 4-bit weight, 6-bit activation precision.

  10. STARFlow: Scaling Latent Normalizing Flows for High-resolution Image Synthesis

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A latent-space transformer autoregressive flow with one deep block plus shallow refiners, tuned noise injection, and score-based guidance reaches competitive FID in high-resolution image synthesis, the first at this s...

  11. Learning on the Manifold: Unlocking Standard Diffusion Transformers with Representation Encoders

    cs.LG 2026-02 conditional novelty 5.0 of 10

    Training flow matching along sphere geodesics with a curvature-aware loss weight lets standard DiT-B converge on DINOv2 features (FID 3.37 with guidance), contradicting the need for width scaling.

Pith tools