method · first seen 4 Oct 2026
Koopman operator learning
Also: Koopman model, Koopman theory, Koopman-inspired, Koopman learning, Koopman operator, Koopman generator, Koopman embeddings, Koopman-inspired model
Learning finite-dimensional linear embeddings of nonlinear dynamics from data, enabling linear analysis of the system.
Koopman operator learning approximates nonlinear dynamics as linear evolution in a learned space of observables, so that linear prediction, spectral analysis and control tools can be applied. The sources show it being used across many applied settings rather than a single new idea: a radar-based ECG recovery framework, a benchmark of Koopman-type representations for latent data assimilation, a NeurIPS paper and accompanying repo on Koopman generative operators for probabilistic forecasting, Koopman-based MPC and model-based reinforcement learning for robotics, calibrated forecasting with Koopman mode decomposition, transition path sampling in molecular dynamics, and battery estimation. The spread is broad but steady, coming mostly from academic papers.
Koopman methods are a long-standing framework (dynamic mode decomposition, deep Koopman autoencoders) whose current use in world models, generative forecasting and control continues an existing line of research rather than marking a new trend. Ongoing work on probabilistic and control-oriented variants may still make linear latent dynamics a more common default for physical-system modeling.
Analyst note (established), 17 Nov 2026.
Mentions per week
PapersCodeDiscussion
Rank 52 on 14 Nov 2026 · 13 mentions in 28 days · 13 independent groups · papers → code · first seen 6 weeks ago · acceleration +1.82
Earliest sources
- Kapture: Capturing Cardiac Dynamics with Koopman-Governed Learning for Efficient Radar-Based Electrocardiogram RecoveryPapers · 4 Oct 2026 · Tong Wu, Jing Peng, Ziqi Feng, Yuanyuan Zhang
- Delay-coordinate reconstruction and conditional-moment causal diagnostics in stochastic systemsPapers · 4 Oct 2026 · Jun Ohkubo
- A Response Theory Probe for Learned Stochastic AI Simulators, Tested on Lorenz-63Papers · 5 Oct 2026 · João Böger, Simon Driscoll, Niccolò Zagli, Valerio Lucarini, Francisco Camara…
- Transition Path Sampling Using Koopman Operators and Exit-Time Optimal ControlPapers · 7 Oct 2026 · Boya Hou, Shane Wang, Siddharth Ambekar, Maxim Raginsky, Olgica Milenkovic
- Koopman Observers for Diffusion Acceleration: Correcting Feature Forecasts with Shallow MeasurementsPapers · 7 Oct 2026 · Hanru Bai, Yuanchao Xu, Fengyi Li
- Understanding Latent-Dimension Scaling in Dynamical-System Learning through Spectral ReliabilityPapers · 8 Oct 2026 · Itsushi Sakata, Yuta Miyauchi, Yoshinobu Kawahara
Latest sources
- SLiDE: Structured Linear Dynamics for Forecasting with Exogenous InputsPapers · 14 Nov 2026 · Sebastian Pütz, Theodore Glavas, Benjamin Schäfer, Boris N. Oreshkin, Mark Coa…
- Prediction-Error-Tightened MPC for Socially Compliant Crowd NavigationPapers · 11 Nov 2026 · Alessandro Febbraro, Federico Rollo, Marcello Farina, Luca Bascetta, Navvab Kash…
- Koopman Model-Based Reinforcement LearningPapers · 10 Nov 2026 · Wenjian Hao, Paulo C. Heredia, Shaoshuai Mou
- K-PWM: Control-Oriented Structured Bilinear World Models under Partial ObservationPapers · 8 Nov 2026 · Santosh Mohan Rajkumar, Sriram Narayanan, Samuel E. Otto, Debdipta Goswami
- Learning calibrated probabilistic forecasts using Koopman mode decomposition and the continuous ranked probability scorePapers · 8 Nov 2026 · Benjamin M. Kent, Richard E. Turner
- Floquet-Parameterized Koopman-MPC for Tracking Periodic OrbitsPapers · 8 Nov 2026 · Kartik A. Pant, Vishnu Vijay, Yufei Zhu, Xintong Deng, Taewoo Nam, Inseok Hwang
- STEK: Structural-Terrain-Enhanced Koopman-Inspired Cross-Channel Regularization for Retrospective Battery Multi-Target EstimationPapers · 4 Nov 2026 · Jing Wang, Shiyuan Piao, Fugee Tsung
- azencot-group/KGOCode · 3 Nov 2026 · azencot-group · 61 stars
- Koopman Generative Operators for Efficient Probabilistic Time-Series ForecastingPapers · 3 Nov 2026 · Raz Marshanski, Liran Nochumsohn, Mayank Jauhari, Boris N. Oreshkin, Omri Azenco…
- When Latent Dynamics Meet Data Assimilation: A Benchmark of Koopman-Type RepresentationsPapers · 2 Nov 2026 · Zhiwei Zhang, Wenxuan Yuan, Xiaoyuan Cheng, Pouria Behnoudfar, Chuanqi Chen, Cha…
- Why Koopman Structure Does Not Help Differentiable World Models for Humanoid Whole-Body TrackingPapers · 31 Oct 2026 · Haizhou Ge, Mengshuo Wang, Yufei Jia, Lei Han, Lu Shi, Guyue Zhou, Ruqi Huang
- Test-Time Prompt-Agnostic DecompositionPapers · 26 Oct 2026 · Junze Wang, Lei Fan, Dezheng Zhang, Donglin Di, Yang Song, Sidong Liu, Cong Cong