Invertible Neural Networks

Follow this topic Ask about this topic

Momentum

1 paper in the last four weeks, with none the four weeks before. 0.0% of all new papers.

Jul 13Week of Sep 28

Latest papers 11

All topics

LLN: Learnable Lens Networks for Parameter-Efficient Long-Horizon Dynamical Prediction
Sep 28, 2026Binbin Yong, Zhao Su, Lan Guo +3Invertible Neural Networks Residual Learning
Lanzhou University · Monash University · University of Wollongong
Conditional Invertible Neural Networks for Data-Driven UAV Control: A 2-D Proof of Concept
Jul 15, 2026Christian Wittke, Stephan Myschik, Oliver NiggemannInvertible Neural Networks Quadrotor Control
Computer Science in Mechanical Engineering, Helmut Schmidt University, Hamburg, Germany. · Institute for Aeronautical Engineering, University of the Bundeswehr Munich, Bavaria, Germany
Robustly Invertible Nonlinear Dynamics and the BiLipREN: From Inversion-Based Control to Generative Trajectory Modelling
Jul 10, 2026Yurui Zhang, Ruigang Wang, Ian R. ManchesterInvertible Neural Networks Recurrent Neural Networks
Australian Centre for Robotics (ACFR), and the School of Aerospace, Mechanical and Mechatronic Engineering, The University of Sydney, Australia
Efficient Cross-Scale Invertible Hiding Network with Spatial-Frequency Collaboration and Non-Invertible Mechanism
Jun 24, 2026Junxue Yang, Xin LiaoInvertible Neural Networks
College of Computer Science and Electronic Engineering, Hunan University, Changsha, China
Invertible Neural Network Adapter for One-Step Flow Matching in Robot Manipulation
Jun 17, 2026Yu Zhang, Kangyi Ji, Yongxiang Zou +3Invertible Neural Networks Flow Matching
High-Fidelity Video Compression based on Invertible Neural Transform and Implicit Conditioning
Jun 11, 2026Siyue Teng, Ho Man Kwan, Yuxuan Jiang +2Invertible Neural Networks Neural Video Compression
Visual Information Lab, University of Bristol, UK
Beyond Linear Activation Steering: Invertible Latent Transformations for Controlling LLM Behavior
Jun 7, 2026Tuc Nguyen, Thai LeInvertible Neural Networks Language Model Steering
Indiana University Bloomington, USA
TVRN: Invertible Neural Networks for Compression-Aware Temporal Video Rescaling
May 15, 2026Xinmin Feng, Li Li, Dong Liu +1Invertible Neural Networks Video Compression
MOE Key Laboratory of Brain-Inspired Intelligent Perception and Cognition, University of Science and Technology of China, Hefei 230093, China
Generative Design of a Gas Turbine Combustor Using Invertible Neural Networks
Apr 27, 2026Patrick Krüger, Hanno Gottschalk, Werner Krebs +1Invertible Neural Networks Inverse Design
Department of Mathematics, Technische Universität Berlin, Straße des 17. Juni 135, Berlin, Berlin, Germany · Sienmens Energy, Rheinstraße 100, Mühlheim an der Ruhr, NRW, Germany
Reversible Residual Normalization Alleviates Spatio-Temporal Distribution Shift
Apr 17, 2026Zhaobo Hu, Vincent Gauthier, Mehdi NaimaInvertible Neural Networks Distribution Shift Robustness
SAMOVAR, Télécom SudParis · Institut Polytechnique de Paris · CNRS – LIP6 +1
Invertible Query-Key Coupling Composes with Attention Mechanisms
Apr 2, 2026Barak Gahtan, Alex M. BronsteinInvertible Neural Networks Self-Attention
Technion – Israel Institute of Technology · ISTA – Institute of Science and Technology Austria

Sep 28, 2026, cs.LGJ/K move · Enter open · S save

LLN: Learnable Lens Networks for Parameter-Efficient Long-Horizon Dynamical Prediction

Binbin Yong, Zhao Su, Lan Guo, Haoran Li+2

Lanzhou University · Monash University · University of Wollongong

Invertible Neural Networks Residual Learning Neural Network Training Dynamics

Open paper AskPDF arXiv

Explicit residual connections of the form (x+f(x)), often combined with normalization layers, have become a standard strategy for training very deep neural networks. However, residual addition primarily provides an algebraic shortcut for gradient propagation, while leaving the evolution of feature geometry across layers largely unconstrained. We introduce Learnable Lens Networks (LLN), a physics-inspired architecture that replaces direct feature-space residual accumulation with learnable optical transport in an augmented position-angle phase space. Each layer alternates between free propagation, which provides an implicit transport path, and a learnable lens field that performs nonlinear trajectory transformation and focusing. Theoretically, we establish that LLN transport is globally invertible and volume-preserving for any differentiable lens field, with the implemented coordinate-wise Gaussian transport further satisfying symplecticity. Importantly, these structural constraints do not limit expressivity: with unrestricted embeddings and readouts, LLN retain universal approximation of continuous end-to-end maps. Experiments across diverse dynamical systems demonstrate that LLN improves long-horizon prediction while using substantially fewer parameters than same-depth comparators. Further analysis reveals stable depth-wise gradient transport and interpretable learned dynamics under the coupled propagation and refraction design.

Momentum

Related topics

Latest papers 11