Generative Adversarial Networks and Image Synthesis

Últimos artículos

• From Core to Detail: Unsupervised Disentanglement with Entropy-Ordered Flows

  Daniel Galperin, Ullrich K\"othe · 9 de febrero de 2026

  Learning unsupervised representations that are both semantically meaningful and stable across runs remains a central challenge in modern representation learning. We introduce entropy-ordered flows (EOFlows), a normalizing-flow framework that orders latent dimensions by their explained entropy, analo…

• Learning a Generative Meta-Model of LLM Activations

  Grace Luo, Jiahai Feng, Trevor Darrell, Alec Radford, Jacob Steinhardt · 9 de febrero de 2026

  Existing approaches for analyzing neural network activations, such as PCA and sparse autoencoders, rely on strong structural assumptions. Generative models offer an alternative: they can uncover structure without such assumptions and act as priors that improve intervention fidelity. We explore this …

• A Lightweight Library for Energy-Based Joint-Embedding Predictive Architectures

  Basile Terver, Randall Balestriero, Megi Dervishi, David Fan, Quentin Garrido, Tushar Nagarajan, Koustuv Sinha, Wancong Zhang, Mike Rabbat, Yann LeCun, Amir Bar · 9 de febrero de 2026

  We present EB-JEPA, an open-source library for learning representations and world models using Joint-Embedding Predictive Architectures (JEPAs). JEPAs learn to predict in representation space rather than pixel space, avoiding the pitfalls of generative modeling while capturing semantically meaningfu…

• RAIGen: Rare Attribute Identification in Text-to-Image Generative Models

  Silpa Vadakkeeveetil Sreelatha, Dan Wang, Serge Belongie, Muhammad Awais, Anjan Dutta · 9 de febrero de 2026

  Text-to-image diffusion models achieve impressive generation quality but inherit and amplify training-data biases, skewing coverage of semantic attributes. Prior work addresses this in two ways. Closed-set approaches mitigate biases in predefined fairness categories (e.g., gender, race), assuming so…

• DAVE: Distribution-aware Attribution via ViT Gradient Decomposition

  Adam Wr\'obel, Siddhartha Gairola, Jacek Tabor, Bernt Schiele, Bartosz Zieli\'nski, Dawid Rymarczyk · 9 de febrero de 2026

  Vision Transformers (ViTs) have become a dominant architecture in computer vision, yet producing stable and high-resolution attribution maps for these models remains challenging. Architectural components such as patch embeddings and attention routing often introduce structured artifacts in pixel-lev…

• Disentanglement by means of action-induced representations

  Gorka Mu\~noz-Gil, Hendrik Poulsen Nautrup, Arunava Majumder, Paulin de Schoulepnikoff, Florian F\"urrutter, Marius Krumm, Hans J. Briegel · 9 de febrero de 2026

  Learning interpretable representations with variational autoencoders (VAEs) is a major goal of representation learning. The main challenge lies in obtaining disentangled representations, where each latent dimension corresponds to a distinct generative factor. This difficulty is fundamentally tied to…

• ReflexFlow: Rethinking Learning Objective for Exposure Bias Alleviation in Flow Matching

  Guanbo Huang, Jingjia Mao, Fanding Huang, Fengkai Liu, Xiangyang Luo, Yaoyuan Liang, Jiasheng Lu, Xiaoe Wang, Pei Liu, Ruiliu Fu, Shao-Lun Huang · 9 de febrero de 2026

  Despite tremendous recent progress, Flow Matching methods still suffer from exposure bias due to discrepancies in training and inference. This paper investigates the root causes of exposure bias in Flow Matching, including: (1) the model lacks generalization to biased inputs during training, and (2)…

• Physics vs Distributions: Pareto Optimal Flow Matching with Physics Constraints

  Giacomo Baldan, Qiang Liu, Alberto Guardone, Nils Thuerey · 9 de febrero de 2026

  Physics-constrained generative modeling aims to produce high-dimensional samples that are both physically consistent and distributionally accurate, a task that remains challenging due to often conflicting optimization objectives. Recent advances in flow matching and diffusion models have enabled eff…

• Di3PO -- Diptych Diffusion DPO for Targeted Improvements in Image

  Sanjana Reddy (Google), Ishaan Malhi (Google DeepMind), Sally Ma (Google DeepMind), Praneet Dutta (Google DeepMind) · 9 de febrero de 2026

  Existing methods for preference tuning of text-to-image (T2I) diffusion models often rely on computationally expensive generation steps to create positive and negative pairs of images. These approaches frequently yield training pairs that either lack meaningful differences, are expensive to sample a…

• Stop the Flip-Flop: Context-Preserving Verification for Fast Revocable Diffusion Decoding

  Yanzheng Xiang, Lan Wei, Yizhen Yao, Qinglin Zhu, Hanqi Yan, Chen Jin, Philip Alexander Teare, Dandan Zhang, Lin Gui, Amrutha Saseendran, Yulan He · 9 de febrero de 2026

  Parallel diffusion decoding can accelerate diffusion language model inference by unmasking multiple tokens per step, but aggressive parallelism often harms quality. Revocable decoding mitigates this by rechecking earlier tokens, yet we observe that existing verification schemes frequently trigger fl…

• Improved Sampling Schedules for Discrete Diffusion Models

  Alberto Foresti, Mustapha Bounoua, Giulio Franzese, Luca Ambrogioni, Pietro Michiardi · 9 de febrero de 2026

  Discrete diffusion models have emerged as a powerful paradigm for generative modeling on sequence data; however, the information-theoretic principles governing their reverse processes remain significantly less understood than those of their continuous counterparts. In this work, we bridge this gap b…

• POP: Online Structural Pruning Enables Efficient Inference of Large Foundation Models

  Yi Chen, Wonjin Shin, Shuhong Liu, Tho Mai, Jeongmo Lee, Chuanbo Hua, Kun Wang, Jun Liu, Joo-Young Kim · 9 de febrero de 2026

  Large foundation models (LFMs) achieve strong performance through scaling, yet current structural pruning methods derive fixed pruning decisions during inference, overlooking sparsity patterns that emerge in the autoregressive token generation. In this paper, we propose POP (Partition-guided Online …

• Latent Structure Emergence in Diffusion Models via Confidence-Based Filtering

  Wei Wei, Yizhou Zeng, Kuntian Chen, Sophie Langer, Mariia Seleznova, Hung-Hsu Chou · 9 de febrero de 2026

  Diffusion models rely on a high-dimensional latent space of initial noise seeds, yet it remains unclear whether this space contains sufficient structure to predict properties of the generated samples, such as their classes. In this work, we investigate the emergence of latent structure through the l…

• Fine-Grained Model Merging via Modular Expert Recombination

  Haiyun Qiu, Xingyu Wu, Liang Feng, Kay Chen Tan · 9 de febrero de 2026

  Model merging constructs versatile models by integrating task-specific models without requiring labeled data or expensive joint retraining. Although recent methods improve adaptability to heterogeneous tasks by generating customized merged models for each instance, they face two critical limitations…

• Rethinking Multi-Condition DiTs: Eliminating Redundant Attention via Position-Alignment and Keyword-Scoping

  Chao Zhou, Tianyi Wei, Yiling Chen, Wenbo Zhou, Nenghai Yu · 9 de febrero de 2026

  While modern text-to-image models excel at prompt-based generation, they often lack the fine-grained control necessary for specific user requirements like spatial layouts or subject appearances. Multi-condition control addresses this, yet its integration into Diffusion Transformers (DiTs) is bottlen…

• NanoFLUX: Distillation-Driven Compression of Large Text-to-Image Generation Models for Mobile Devices

  Ruchika Chavhan, Malcolm Chadwick, Alberto Gil Couto Pimentel Ramos, Luca Morreale, Mehdi Noroozi, Abhinav Mehrotra · 9 de febrero de 2026

  While large-scale text-to-image diffusion models continue to improve in visual quality, their increasing scale has widened the gap between state-of-the-art models and on-device solutions. To address this gap, we introduce NanoFLUX, a 2.4B text-to-image flow-matching model distilled from 17B FLUX.1-S…

• Diffusion Model's Generalization Can Be Characterized by Inductive Biases toward a Data-Dependent Ridge Manifold

  Ye He, Yitong Qiu, Molei Tao · 6 de febrero de 2026

  When a diffusion model is not memorizing the training data set, how does it generalize exactly? A quantitative understanding of the distribution it generates would be beneficial to, for example, an assessment of the model's performance for downstream applications. We thus explicitly characterize wha…

• Alignment of Diffusion Models: Fundamentals, Challenges, and Future

  Buhua Liu, Shitong Shao, Bao Li, Lichen Bai, Zhiqiang Xu, Haoyi Xiong, James Kwok, Sumi Helal, Zeke Xie · 6 de febrero de 2026

  Diffusion models have emerged as the leading paradigm in generative modeling, excelling in various applications. Despite their success, these models often misalign with human intentions and generate results with undesired properties or even harmful content. Inspired by the success and popularity of …

• Variational Speculative Decoding: Rethinking Draft Training from Token Likelihood to Sequence Acceptance

  Xiandong Zou, Jianshu Li, Jing Huang, Pan Zhou · 6 de febrero de 2026

  Speculative decoding accelerates inference for (M)LLMs, yet a training-decoding discrepancy persists: while existing methods optimize single greedy trajectories, decoding involves verifying and ranking multiple sampled draft paths. We propose Variational Speculative Decoding (VSD), formulating draft…

• Causal Representation Meets Stochastic Modeling under Generic Geometry

  Jiaxu Ren, Yixin Wang, Biwei Huang · 6 de febrero de 2026

  Learning meaningful causal representations from observations has emerged as a crucial task for facilitating machine learning applications and driving scientific discoveries in fields such as climate science, biology, and physics. This process involves disentangling high-level latent variables and th…

• FlashBlock: Attention Caching for Efficient Long-Context Block Diffusion

  Zhuokun Chen, Jianfei Cai, Bohan Zhuang · 6 de febrero de 2026

  Generating long-form content, such as minute-long videos and extended texts, is increasingly important for modern generative models. Block diffusion improves inference efficiency via KV caching and block-wise causal inference and has been widely adopted in diffusion language models and video generat…

• SurvDiff: A Diffusion Model for Generating Synthetic Data in Survival Analysis

  Marie Brockschmidt, Maresa Schr\"oder, Stefan Feuerriegel · 6 de febrero de 2026

  Survival analysis is a cornerstone of clinical research by modeling time-to-event outcomes such as metastasis, disease relapse, or patient death. Unlike standard tabular data, survival data often come with incomplete event information due to dropout, or loss to follow-up. This poses unique challenge…

• Temporal Pair Consistency for Variance-Reduced Flow Matching

  Chika Maduabuchi, Jindong Wang · 6 de febrero de 2026

  Continuous-time generative models, such as diffusion models, flow matching, and rectified flow, learn time-dependent vector fields but are typically trained with objectives that treat timesteps independently, leading to high estimator variance and inefficient sampling. Prior approaches mitigate this…

• DisCa: Accelerating Video Diffusion Transformers with Distillation-Compatible Learnable Feature Caching

  Chang Zou, Changlin Li, Yang Li, Patrol Li, Jianbing Wu, Xiao He, Songtao Liu, Zhao Zhong, Kailin Huang, Linfeng Zhang · 6 de febrero de 2026

  While diffusion models have achieved great success in the field of video generation, this progress is accompanied by a rapidly escalating computational burden. Among the existing acceleration methods, Feature Caching is popular due to its training-free property and considerable speedup performance, …

• Orthogonal Model Merging

  Sihan Yang, Kexuan Shi, Weiyang Liu · 6 de febrero de 2026

  Merging finetuned Large Language Models (LLMs) has become increasingly important for integrating diverse capabilities into a single unified model. However, prevailing model merging methods rely on linear arithmetic in Euclidean space, which often destroys the intrinsic geometric properties of pretra…