物理约束 ML论文趋势 2023–2026

主要结论

• 物理约束 ML论文从 2023 年的 50 篇增长到 2026 年的 199 篇，三年间扩大到约 4.0 倍。
• NeurIPS 是发物理约束 ML论文最多的会议，2023–2026 年共 167 篇，占四个会议总量的 35%。
• 交叉方向里涨得最快的是多模态大模型：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 11 篇。
• 在「气候与物理科学」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的天气预报是 6.1 倍，物理约束 ML是 4.0 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026Derivative-Informed Training of Neural Operators On-the-Fly via Sketched Tangent Consistency
• ICML 2026Bias-Spectrum Neural Processes for Parametric PDEs: Architecture Priors Meet PDE Constraints
• ICLR 2026CONSIGN: Conformal Segmentation Informed by Spatial Groupings via Decomposition
• CVPR 2026Domain Sensitive Federated Learning with Fisher-Informed Pruning
• NeurIPS 2025Activation-Informed Merging of Large Language Models
• ICML 2025Geometry-Informed Neural Networks
• ICLR 2025Beyond the convexity assumption: Realistic tabular data generation under quantifier-free real linear constraints
• CVPR 2025A Physics-Informed Blur Learning Framework for Imaging Systems
• NeurIPS 2024ChaosBench: A Multi-Channel, Physics-Based Benchmark for Subseasonal-to-Seasonal Climate Prediction
• ICML 2024A Space Group Symmetry Informed Network for O(3) Equivariant Crystal Tensor Prediction
• ICLR 2024AirPhyNet: Harnessing Physics-Guided Neural Networks for Air Quality Prediction
• CVPR 2024A Physics-informed Low-rank Deep Neural Network for Blind and Universal Lens Aberration Correction
• NeurIPS 2023ClimSim: A large multi-scale dataset for hybrid physics-ML climate emulation
• ICML 2023Gradient Descent Finds the Global Optima of Two-Layer Physics-Informed Neural Networks
• ICLR 2023Bi-level Physics-Informed Neural Networks for PDE Constrained Optimization using Broyden's Hypergradients
• CVPR 20233D Human Pose Estimation via Intuitive Physics
• NeurIPS 2026Neural Modal Decomposition: Architectural Priors from Observables
• ICML 2026Fine-to-Coarse Fairness-Informed Multi-View Clustering
• ICLR 2026From Neural Networks to Logical Theories: The Correspondence between Fibring Modal Logics and Fibring Neural Networks
• CVPR 2026LaDy Lagrangian-Dynamic Informed Network for Skeleton-based Action Segmentation via Spatial-Temporal

全部物理约束 ML论文

共 167 篇，第 1 / 4 页

1. NeurIPS 2026AnyBokeh: Physics-Guided Any-to-Any Bokeh Editing with Optical Fingerprint Transfer
2. NeurIPS 2026Architecture-Embedded Physics Priors for Mitigating Spectral Bias in Physics-Informed Neural Networks
3. NeurIPS 2026Context Value Informed In-Context Reinforcement Learning
4. NeurIPS 2026CrystalREPA: Transferring Physical Priors from Universal MLIPs to Crystal Generative Models
5. NeurIPS 2026Derivative-Informed Training of Neural Operators On-the-Fly via Sketched Tangent Consistency
6. NeurIPS 2026Diffeomoprhism-Informed 3D Gaussian Splattings via Screen-Space Optimal Transport (OT)
7. NeurIPS 2026End-to-End Differentiable Diffusion Conditioning for Physics-Informed Optimization
8. NeurIPS 2026Energy-Guided Transport for Projection-Free Physics-Informed Flow Matching
9. NeurIPS 2026Fast Learning Rates for Physics-Informed Kernel Methods
10. NeurIPS 2026Generalizable Physics Simulation through Compositional Energy Minimization
11. NeurIPS 2026Geometry Meets Physics: Data-Efficient Pre-Training for Unstructured Neural PDE Solvers
12. NeurIPS 2026GIVLA: Deep Geometry Internalization for A Lightweight VLA via Geometry Instruction and Gradient-Informed Training
13. NeurIPS 2026H-Flow: Self-supervised Human Scene Flow via Physics-inspired Joint Multi-modal Learning
14. NeurIPS 2026INEUS: Iterative Neural Solver for High-Dimensional PIDEs
15. NeurIPS 2026Informed Posterior Sampling: More Efficient Online Learning with Few Offline Demonstrations in Average-Reward MDPs
16. NeurIPS 2026KerONet: A Single Softmax Readout Suffices for Physics-Informed Operator Learning
17. NeurIPS 2026Learning POMDP World Models from Observations with Language-Model Priors
18. NeurIPS 2026LEMON-ZEST: Evolution-Informed Tokenization for Efficient Protein Language Modeling
19. NeurIPS 2026Mosaic: A Benchmark Suite for Differentiable Physics Solvers
20. NeurIPS 2026Neural Field Thermal Tomography: A Differentiable Physics Framework for Non-Destructive Evaluation
21. NeurIPS 2026Neural Modal Decomposition: Architectural Priors from Observables
22. NeurIPS 2026Neural Operator-based Curriculum Learning for Physics-Informed Neural Networks
23. NeurIPS 2026NeuralFieldManifold: Reconstruction of LFP manifold with Lag Embedding
24. NeurIPS 2026Newton-PINet: A fast physics-informed neural network with Newton linearization for meta-learning nonlinear PDEs
25. NeurIPS 2026PDE-PFN: Prior-Data Fitted Neural PDE Solver
26. NeurIPS 2026Per-Loss Adapters for Gradient Conflict in Physics-Informed Neural Networks
27. NeurIPS 2026PhysElite: How Far Are LLMs from Solving Olympiad-Level Physics Problems?
28. NeurIPS 2026PhysFormer: Learning to Simulate Mechanics in World Space
29. NeurIPS 2026Physics Unrolled Neural Operator for Wireless Field Modeling
30. NeurIPS 2026Physics-Conditioned Video Diffusion with Kinematic Priors for Fusion Capsule Polishing
31. NeurIPS 2026Physics-Informed Functional Tucker Method with RKHS Factors for Sparse Spatiotemporal Reconstruction
32. NeurIPS 2026Physics-Informed Optimal Control by Control-Only Supervision with Error Guarantees on Value and Policy
33. NeurIPS 2026Physics-Informed Video Generation via Mixture-of-Experts Latent Alignment
34. NeurIPS 2026PhySPRING: Structure-Preserving Reduction of Physics-Informed Digital Twins via Graph Neural Networks
35. NeurIPS 2026PI-EDG: Physics-Informed Full-Space Electron Density Generation from Molecular Geometry
36. NeurIPS 2026PINNeval: A Comprehensive Evaluation Standard for Physics-Informed Neural Networks
37. NeurIPS 2026PIU-CR: Physics-Informed Deep Unfolding Network with SAR-Optical Image Fusion for Cloud Removal
38. NeurIPS 2026PolarScale: A Physics-Grounded Benchmark for Radiometrically Consistent RGB-to-Stokes Estimation
39. NeurIPS 2026Prior Text-Informed Gate Attention Framework for Multimodal Psychiatric Disorder Diagnosis
40. NeurIPS 2026RAWild: Toward Sensor-Agnostic RAW Object Detection via Physics-Guided Curve and Grid Modeling
41. NeurIPS 2026ReGuidance: Diffusion Steering with Strong Latent Initializations Solves Hard Inverse Problems
42. NeurIPS 2026REINS: Learning Inertia-Induced Geometry for Physics-Consistent Motion Representation in Clinical Gait Phenotyping
43. NeurIPS 2026Self-Supervised Doppler-Guided RF Odometry
44. NeurIPS 2026Simplifying Transformer-Based U-Net Neural Physics Simulators
45. NeurIPS 2026SKIM: Pruning Large Language Model Agents via Selective Knowledge Informed Masking
46. NeurIPS 2026Sparsely-Supervised Data Assimilation via Physics-Informed Schrödinger Bridge
47. NeurIPS 2026SpikeSSL: A Universal Spike Inference Framework with Dynamics-Informed State-Space Layers
48. NeurIPS 2026The Subjectivity of Monoculture
49. NeurIPS 2026Training-Free Cultural Alignment of Large Language Models via Persona Disagreement
50. NeurIPS 2026Ultra Fast PDE Solving via Physics Guided Few-step Diffusion

常见问题

2023–2026 年四大会议一共有多少篇物理约束 ML论文？

共 481 篇，按年份分别是 2023 年 50 篇、2024 年 85 篇、2025 年 147 篇、2026 年 199 篇。

NeurIPS 2026 有多少篇物理约束 ML论文？

NeurIPS 2026 有 54 篇物理约束 ML论文，上一届是 65 篇。

ICLR 2026 有多少篇物理约束 ML论文？

ICLR 2026 有 64 篇物理约束 ML论文，上一届是 38 篇。

ICML 2026 有多少篇物理约束 ML论文？

ICML 2026 有 51 篇物理约束 ML论文，上一届是 25 篇。

CVPR 2026 有多少篇物理约束 ML论文？

CVPR 2026 有 30 篇物理约束 ML论文，上一届是 19 篇。

哪个会议的物理约束 ML论文最多？

NeurIPS，2023–2026 年共 167 篇。