Deep networks often rely on architectural heuristics to shape representation evolution, limiting their ability to model data governed by intrinsic dynamics. We present the Circuit-inspired High-Order Neural Network (CHONN), a modular framework that treats representation evolution as a latent potential process and increases its effective order through Kirchhoff-inspired cascade composition. A single Kirchhoff Neural Cell implements a stable first-order update, while serially composed cells form higher-order dynamical operators within one block. This construction is interpretable, numerically stable and compatible with common neural backbones. Theoretical analysis shows that cascaded cells induce end-to-end high-order operators, and controlled experiments demonstrate that intra-block high-order construction differs from generic depth stacking, especially on derivative-sensitive measures. Across steady-state operator learning, long-horizon physical forecasting and ImageNet-1K recognition, CHONN improves structural fidelity, rollout stability and visual representation learning. These results identify high-order circuit composition as a general principle for neural dynamics modeling.
SP2RINT: Spatially-Decoupled Physics-Inspired Progressive Inverse Optimization for Scalable, PDE-Constrained Meta-Optical Neural Network Training
School of Electrical, Computer and Energy Engineering, Arizona State University（电气、计算机与能源工程学院，亚利桑那州立大学）；Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology（电气工程与计算机科学系，麻省理工学院）；Department of Electrical, Computer, and System Engineering, Rensselaer Polytechnic Institute（电气、计算机与系统工程系，拉特格斯理工学院）；Department of Electrical and Computer Engineering, University of Wisconsin at Madison（电气与计算机工程系，威斯康星大学麦迪逊分校）
Probing neural representations of scene perception in a hippocampally dependent task using artificial neural networks
Kavli Institute for Systems Neuroscience（卡夫利系统神经科学研究所）；NTNU（挪威科技大学）；Max-Planck-Insitute for Human Cognitive and Brain Sciences（马克斯·普朗克人类认知与脑科学研究所）；UCL（伦敦大学学院）
Hypergraph Neural Diffusion: A PDE-Inspired Framework for Hypergraph Message Passing
School of Mathematics and Statistics, Shandong University（山东大学数学与统计学院）；Academy of Mathematics and Systems Science, Chinese Academy of Sciences（中国科学院数学与系统科学研究院）；School of Mathematical Sciences, University of Chinese Academy of Sciences（中国科学院大学数学科学学院）；Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）；School of Cyber Security, University of Chinese Academy of Sciences（中国科学院大学网络空间安全学院）
Motion Inspired Unsupervised Perception and Prediction in Autonomous Driving
Waymo LLC（Waymo有限责任公司）
现代自动驾驶系统中基于学习的感知和预测模块通常依赖于昂贵的人工注释，并且被设计为仅感知少数预定义的对象类别。这种闭集范式不足以满足安全关键的自动驾驶任务，其中自动驾驶车辆需要在高度动态的世界中处理任意多种类型的交通参与者及其运动行为。为了解决这个困难，本文开创了一个新颖而具有挑战性的方向，即，训练感知和预测模型，以理解开放集移动对象，无需人工监督。我们提出的框架使用自学习流来触发自动Meta标记管道，以实现自动监督。Waymo Open Dataset上的3D检测实验表明，我们的方法明显优于经典的无监督方法，甚至与有监督场景流的方法相比具有竞争力。我们进一步表明，我们的方法在开集3D检测和轨迹预测方面产生了非常有希望的结果，证实了它在缩小完全监督系统的安全差距方面的潜力。
Beyond Domain Randomization: Event-Inspired Perception for Visually Robust Adversarial Imitation from Videos
Mercedes-Benz AG（梅赛德斯-奔驰集团）；Technical University of Munich（慕尼黑技术大学）；Purdue University（普渡大学）；Space Robotics Research Group, SnT, University of Luxembourg（卢森堡大学空间机器人研究组）；Torc Robotics（Torc机器人公司）；Princeton University（普林斯顿大学）
LaPON: A Lagrange's-mean-value-theorem-inspired operator network for solving PDEs and its application on NSE
Ocean College, Zhejiang University（浙江大学海洋学院）；Ocean Research Center of Zhoushan, Zhejiang University（浙江大学舟山海洋研究中心）；Hangzhou International Innovation Institute, Beihang University（北京航空航天大学杭州国际创新研究院）
Self-composing neural operators for high-frequency and multiscale PDE surrogates
Yau Mathematical Sciences Center, Tsinghua University（清华大学姚期智数学科学中心）；Ocean University of China（中国海洋大学）；King Abdullah University of Science and Technology（国王 Abdullah 科学与技术大学）
NODE-ImgNet: a PDE-informed effective and robust model for image denoising
University of Strathclyde（斯特拉斯克莱德大学）；University College London（伦敦大学学院）；The Alan Turing Institute（艾伦·图灵研究所）；Oklahoma State University（俄克拉荷马州立大学）；University of Georgia（佐治亚大学）
AOT-POT: Adaptive Operator Transformation for Large-Scale PDE Pre-training
University of Science and Technology of China（中国科学技术大学）；Shanghai AI Laboratory（上海人工智能实验室）；Department of Electronic Engineering, Tsinghua University, China（清华大学电子工程系）
BENO: Boundary-embedded Neural Operators for Elliptic PDEs
National Engineering Research Center for Software Engineering, Peking University（北京大学软件工程国家工程研究中心）；Department of Engineering, Westlake University（西湖大学工学院）；Department of Astronautics and Aeronautics, Stanford University（斯坦福大学航空航天系）
Solving high-dimensional Hamilton-Jacobi-Bellman PDEs using neural networks: perspectives from the theory of controlled diffusions and measures on path space
Universität Potsdam（波茨坦大学）；Freie Universität Berlin（柏林自由大学）；Brandenburgische Technische Universität Cottbus-Senftenberg（勃兰登堡工业大学科特布斯-森夫滕贝格）
School of Mathematical Sciences（数学科学学院）；School of Mathematical Sciences, Peking University（北京大学数学科学学院）；Peking University, Beijing, China（北京大学北京中国）；Beijing Computational Science Research Center（北京计算科学研究中心）；Beijing International Center for Mathematical Research, Peking University（北京大学北京国际数学研究中心）；Center for Data Science, Peking University（北京大学数据科学中心）；Beijing Institute of Big Data Research（北京大数据研究院）
Enhancing Solutions for Complex PDEs: Introducing Complementary Convolution and Equivariant Attention in Fourier Neural Operators
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）
