Image de couverture de livre

Learning to Think in Physics: Breaking Shortcut Learning in Scientific Diffusion via Representation Alignment

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Jia, Haozhe, Yin, Pengyu, Chen, Wenshuo, Liang, Shaofeng, Wang, Lei, Tian, Bowen, Wang, Xiucheng, Jia, Nanqian, Yue, Yutao
Format:	Preprint
Publié:	2026
Sujets:	Machine Learning Computer Vision and Pattern Recognition
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Physics-Informed Representation Alignment for Sparse Radio-Map Reconstruction
par: Jia, Haozhe, et autres
Publié: (2025)

Oracle Noise: Faster Semantic Spherical Alignment for Interpretable Latent Optimization
par: Li, Haosen, et autres
Publié: (2026)

LUMA: Low-Dimension Unified Motion Alignment with Dual-Path Anchoring for Text-to-Motion Diffusion Model
par: Jia, Haozhe, et autres
Publié: (2025)

Guided Path Sampling: Steering Diffusion Models Back on Track with Principled Path Guidance
par: Li, Haosen, et autres
Publié: (2025)

POLARIS: Projection-Orthogonal Least Squares for Robust and Adaptive Inversion in Diffusion Models
par: Chen, Wenshuo, et autres
Publié: (2025)

$Z^2$-Sampling: Zero-Cost Zigzag Trajectories for Semantic Alignment in Diffusion Models
par: Li, Haosen, et autres
Publié: (2026)

Delta Score Matters! Spatial Adaptive Multi Guidance in Diffusion Models
par: Li, Haosen, et autres
Publié: (2026)

RadioFlow: Efficient Radio Map Construction Framework with Flow Matching
par: Jia, Haozhe, et autres
Publié: (2025)

Before the Body Moves: Learning Anticipatory Joint Intent for Language-Conditioned Humanoid Control
par: Jia, Haozhe, et autres
Publié: (2026)

Free-T2M: Robust Text-to-Motion Generation for Humanoid Robots via Frequency-Domain
par: Chen, Wenshuo, et autres
Publié: (2025)

Neural Electromagnetic Fields for High-Resolution Material Parameter Reconstruction
par: Chen, Zhe, et autres
Publié: (2026)

Breaking Scale Anchoring: Frequency Representation Learning for Accurate High-Resolution Inference from Low-Resolution Training
par: Wang, Wenshuo, et autres
Publié: (2025)

ANT: Adaptive Neural Temporal-Aware Text-to-Motion Model
par: Chen, Wenshuo, et autres
Publié: (2025)

AnyEdit++: Adaptive Long-Form Knowledge Editing via Bayesian Surprise
par: Tian, Bowen, et autres
Publié: (2026)

Latent Diffusion Model Based Denoising Receiver for 6G Semantic Communication: From Stochastic Differential Theory to Application
par: Wang, Xiucheng, et autres
Publié: (2025)

ECHO: Edge-Cloud Humanoid Orchestration for Language-to-Motion Control
par: Jia, Haozhe, et autres
Publié: (2026)

Text2Weight: Bridging Natural Language and Neural Network Weight Spaces
par: Tian, Bowen, et autres
Publié: (2025)

Super-Resolution Enhancement of Medical Images Based on Diffusion Model: An Optimization Scheme for Low-Resolution Gastric Images
par: Jia, Haozhe
Publié: (2025)

RadioDiff-FS: Physics-Informed Manifold Alignment in Few-Shot Diffusion Models for High-Fidelity Radio Map Construction
par: Wang, Xiucheng, et autres
Publié: (2026)

RadioMamba: Breaking the Accuracy-Efficiency Trade-off in Radio Map Construction via a Hybrid Mamba-UNet
par: Jia, Honggang, et autres
Publié: (2025)

AlignDiff: Learning Physically-Grounded Camera Alignment via Diffusion
par: Xie, Liuyue, et autres
Publié: (2025)

Improving the Classification Effect of Clinical Images of Diseases for Multi-Source Privacy Protection
par: Bowen, Tian, et autres
Publié: (2024)

Point-SRA: Self-Representation Alignment for 3D Representation Learning
par: Wei, Lintong, et autres
Publié: (2026)

SLAP: Shortcut Learning for Abstract Planning
par: Liu, Y. Isabel, et autres
Publié: (2025)

Conditional Information Bottleneck for Multimodal Fusion: Overcoming Shortcut Learning in Sarcasm Detection
par: Wang, Yihua, et autres
Publié: (2025)

RadioMapMotion: A Dataset and Baseline for Proactive Spatio-Temporal Radio Environment Prediction
par: Jia, Honggang, et autres
Publié: (2025)

Step-level Denoising-time Diffusion Alignment with Multiple Objectives
par: Zhang, Qi, et autres
Publié: (2026)

Mixture of Group Experts for Learning Invariant Representations
par: Kang, Lei, et autres
Publié: (2025)

Model-Free Robust Reinforcement Learning with Sample Complexity Analysis
par: Wang, Yudan, et autres
Publié: (2024)

Wolf2Pack: The AutoFusion Framework for Dynamic Parameter Fusion
par: Tian, Bowen, et autres
Publié: (2024)

Multimodal Representation Alignment for Image Generation: Text-Image Interleaved Control Is Easier Than You Think
par: Chen, Liang, et autres
Publié: (2025)

PointDico: Contrastive 3D Representation Learning Guided by Diffusion Models
par: Li, Pengbo, et autres
Publié: (2025)

Learn for Variation: Variationally Guided AAV Trajectory Learning in Differentiable Environments
par: Wang, Xiucheng, et autres
Publié: (2026)

Reinforcement Learning with Physics-Informed Symbolic Program Priors for Zero-Shot Wireless Indoor Navigation
par: Li, Tao, et autres
Publié: (2025)

ShortFT: Diffusion Model Alignment via Shortcut-based Fine-Tuning
par: Guo, Xiefan, et autres
Publié: (2025)

Demonstrating and Reducing Shortcuts in Vision-Language Representation Learning
par: Bleeker, Maurits, et autres
Publié: (2024)

Achieving the Asymptotically Optimal Sample Complexity of Offline Reinforcement Learning: A DRO-Based Approach
par: Wang, Yue, et autres
Publié: (2023)

Emotion-Director: Bridging Affective Shortcut in Emotion-Oriented Image Generation
par: Jia, Guoli, et autres
Publié: (2025)

Reasoning Physical Video Generation with Diffusion Timestep Tokens via Reinforcement Learning
par: Lin, Wang, et autres
Publié: (2025)

Break the Chain: Large Language Models Can be Shortcut Reasoners
par: Ding, Mengru, et autres
Publié: (2024)