imagenet

Hoover Institution's 'Only in America' interview with Fei-Fei Li — the 'Godmother of AI' — covering her journey to computer science, ImageNet's role in the deep learning revolution, and how World Labs' 'world models' will extend generative AI into spatial intelligence, plus discussions on truth-seeking and U.S. AI preparedness.

分析了连续思维机（CTM）在ImageNet图像分类上的内部动力学，发现其思维轨迹是移动吸引不动点的曲线，注意力调度在图像间基本通用，表明CTM已找到一种处理所有ImageNet图像的统一策略。

本文提出 data-space iteration 框架，将少步生成的迭代过程从概率流空间移至数据空间，从而完全移除对流离散化时间表的依赖，并在 ImageNet 256x256 类条件生成上以最小改动集成 DMD，超越了需要手动调参时间表的基线方法。

14 years after AlexNet's ImageNet submission kicked off the modern deep learning revolution, the author celebrates by showcasing an open-source creative pipeline using Qwen-Image 2.1, LTX-2.3, Wan 2.1, InfiniteTalk, and ComfyUI running on a single RTX 3090.

The paper presents Improved Distributional Diffusion Models (iDDM), which overcome scaling limits of DDMs via late particle expansion and time-dependent scoring rule schedules, achieving 4.48 FID at 4 steps and 2.38 FID at 50 steps on ImageNet-256 with DiT-XL/2 from a single-stage from-scratch training without teachers or self-distillation.

The paper introduces Persistence Forcing (PerF), which assigns heterogeneous refinement budgets to feature groups in pixel-space Diffusion Transformers, producing persistent global-structure features that guide active detail features. PerF-H reaches FID 1.63 on ImageNet 256x256, outperforming JiT-H with competitive efficiency.

This paper introduces Representation Distribution Matching (RDM), a method for one-step image generation by matching feature distributions under pretrained encoders, achieving state-of-the-art results on ImageNet and enabling post-training of FLUX.2 into a one-step generator with improved performance.

Parallel Rollout Approximation (PRA) improves pixel-space autoregressive image generation by using low-dimensional intermediate states and parallel training, achieving new state-of-the-art results on ImageNet-1K generation.

Un-0 is an image generator powered by a simulated system of coupled oscillators, achieving FID 6.74 on ImageNet 64×64, matching early conventional methods. It is open-source and aims to demonstrate energy-efficient AI on physical substrates.

ViT-Up introduces a task-agnostic feature upsampler for Vision Transformers that predicts features at arbitrary continuous image coordinates, enabling dense feature maps at any resolution and improving dense prediction and semantic correspondence benchmarks. It outperforms prior state-of-the-art upsamplers, with gains of up to +2.07 mIoU on Cityscapes and +4.17 [email protected] on SPair-71k.

This paper investigates whether the Engram module, an associative memory mechanism, provides content-addressable retrieval in autoregressive image generation or acts as a gated architectural side-pathway.

This paper proposes aligning latent geometry for spherical flow matching, projecting latents onto a fixed-radius sphere and using spherical linear interpolation to improve image generation quality, consistently improving FID on class-conditional ImageNet.