arXiv Machine Learning

NAE: Normalizing AutoEncoder

arXiv:2608. 12084v1 Announce Type: new Abstract: We consider the setting of Normalizing flows with approximate inverses, an established paradigm spanning both full-dimensional ($d=D$) and bottleneck ($d<D$) settings, and group these models under the term flow autoencoders.

arXiv Machine Learning
Jul 15

SinAE: A Single-Architecture Flow-Matching Autoencoder for Cross-Domain Atomic Systems

arXiv:2607. 12380v1 Announce Type: new Abstract: Small molecules, crystals, and proteins all reduce to atoms in 3D space, yet their generative pipelines remain fragmented across domains, each with its Small molecules, crystals, and proteins all reduce to atoms in 3D space, yet their generative pipelines remain fragmented across domains, each with its own graph, equivariant, or frame-based architecture.

By Yuxuan Ren, Fan Yang, Jianhua Yao, Yatao Bian
arXiv Machine Learning
Jul 21

Routing by Reaching: Composition of Pre-trained GFlowNets for Multi-Objective Generation

arXiv:2602. 21565v3 Announce Type: replace Abstract: Generative Flow Networks (GFlowNets) learn to sample diverse candidates in proportion to a reward function, making them well-suited for scientific discovery, where exploring multiple promising solutions is crucial.

By Seokwon Yoon, Youngbin Choi, Seunghyuk Cho, Seungbeom Lee, MoonJeong Park, Dongwoo Kim
arXiv AI
Aug 12

Flow Straight to Reality: Perceptually Consistent Flow Matching for Efficient Image Restoration

arXiv:2608. 10544v1 Announce Type: cross Abstract: Image restoration is fundamentally constrained by the tradeoff between distortion and perception: minimizing pixel-wise error yields over-smoothed results, whereas optimizing for perceptual realism often introduces structural deviations.

By Sangwoo Jo, Donggeun Ko, Jayeon Kang, Youngsang Kwak, Jaehwa Kwak, Sungjoon Choi
arXiv Machine Learning
Jun 15

Smoothing Dark Areas in Molecular Latent Diffusion

arXiv:2606. 13955v1 Announce Type: new Abstract: Latent diffusion is a promising framework for scalable 3D molecular generation, but it requires a latent space that remains smooth, valid, and navigable beyond posterior samples.

By Xi Wang, Jiahan Li, Yuxuan Xia, Yingcheng Wu, Shaoyi Zheng, Shengjie Wang
Hugging Face Trending Papers
Jun 9

IDEAL: In-DEpth ALignment Makes A Discrete Representation AutoEncoder

Built on pretrained vision foundation models (VFMs), representation autoencoders (RAEs) have recently emerged as a promising approach for constructing semantically rich latent spaces for image generation. However, their reconstruction quality often remains suboptimal, largely because deep VFM representations do not preserve sufficient fine-grained visual detail.