Dynamical Systems

Follow this topic Ask about this topic

Momentum

44 papers in the last four weeks, up 159% on the four weeks before. 0.4% of all new papers.

Jul 13Week of Sep 28

Latest papers 319

All topics

DAGGER: Gradient-Free Construction of Transiently Amplifying Networks under Hard Connectivity Constraints
May 31, 2026James C. FergusonDynamical Systems Sparse Neural Networks
The African Institute for Mathematical Sciences (Muizenberg, South Africa) · the Institute of Science and Technology Austria (Klosterneuburg, Austria)
Tiny Recursive Models for Solving the J2-Perturbed Lambert Problem
May 30, 2026Minduli Wijayatunga, Roberto ArmellinDynamical Systems Trajectory Optimization
Department of Aerospace Engineering University of Illinois Urbana-Champaign Urbana-Champaign, IL, USA · Te Pūnaha Ātea – Space Institute University of Auckland Auckland, New Zealand
The Dynamic-Probabilistic Consistency Gap in Chaotic Surrogate Modeling
May 29, 2026Andre Herz, Matthijs Pals, Daniel Durstewitz +1Dynamical Systems Surrogate Modeling
1Interdisciplinary Center for Scientific Computing, Heidelberg University, Germany · Faculty of Mathematics and Computer Science, Heidelberg University, Germany · 3Dept. of Theoretical Neuroscience, Central Institute of Mental Health (CIMH), Mannheim, Germany +3
LFNO: Bridging Laplace and Fourier via Transient-Steady Decomposition
May 29, 2026Jeongun Ha, Sanga Yoon, Donghun LeeDynamical Systems Neural PDE Solvers
Department of Mathematics, Korea University, 145 Anam-ro Seongbuk-gu Seoul, Republic of Korea.
Learning Permutation-invariant Macroscopic Dynamics
May 29, 2026Zhichao Han, Mengyi Chen, Qianxiao LiDynamical Systems Latent Dynamics Modeling
Institute for Functional Intelligent Materials, National University of Singapore · Department of Mathematics, National University of Singapore
Learning Transferable Predictability Representations
May 28, 2026Diyali Goswami, Auroop R. GangulyDynamical Systems Ordinal Regression
Sustainability and Data Sciences Laboratory (SDS Lab), Northeastern University, Boston, MA, USA · AI4CaS: AI for Climate and Sustainability, Institute for Experiential AI, Northeastern University, Boston, MA, USA · Pacific Northwest National Laboratory (PNNL), Richland, WA, USA
Delayed Repression and Emergent Instability in Adaptive Multi-Agent Systems
May 28, 2026Igor ItkinDynamical Systems Replicator Dynamics
Independent Researcher, Tel Aviv, Israel
Stochastic Lifting for Generating Trajectories of Stochastic Physical Systems
May 28, 2026Jules Berman, Tobias Blickhan, Benjamin PeherstorferDynamical Systems Stochastic Linear Dynamical Systems
Courant Institute of Mathematical Sciences, New York University, New York, NY 10012, USA.
A Minimal Bifurcation Model of Load Imbalance in a Softmax Mixture-of-Experts Router
May 27, 2026O. M. KiselevDynamical Systems Expert Load Balancing
Innopolis University, Innopolis, Russia
A Multiscale Kinetic Framework for Image Segmentation: From Particle Systems to Continuum Models
May 27, 2026Horacio Tettamanti, Giulia Guicciardi, Mattia ZanellaDynamical Systems Image Segmentation
Department of Mathematics “F. Casorati” University of Pavia, Italy
Sampling Data with Chains of Forward-Backward Diffusion Steps
May 26, 2026Hyunmo Kang, Noam Itzhak Levi, Corinna Elena Wegner +2Dynamical Systems Markov Chain Monte Carlo
Johns Hopkins University · EPFL · EPFL & University of Göttingen +1
Two-Parameter Flows for Learning Population Dynamics of Physical Systems
May 25, 2026Paul Schwerdtner, Tobias Blickhan, Benjamin PeherstorferDynamical Systems Continuous Normalizing Flows
Courant Institute of Mathematical Sciences, New York University, 251 Mercer Street, New York, NY 10012, USA.
Emergence via Phase Transitions: Mechanism Landscapes and Universal Convergence Across Complex Systems
May 25, 2026Truong Xuan KhanhDynamical Systems Phase Transitions
H&K Research Studio, Clevix LLC, Hanoi, Vietnam
Modeling Dynamic Mixtures of Time-Delay Systems from Streaming Time Series
May 25, 2026Ren Fujiwara, Yasuko Matsubara, Yasushi SakuraiDynamical Systems Online System Identification
SANKEN, The University of Osaka, Japan
Beyond Fixed Points: Superpolynomial Capacity of Asymmetric Hopfield Networks
May 23, 2026Aakash Kumar, Anatoly Khina, Frederik Mallmann-Trenn +1Dynamical Systems Hopfield Networks
COATI, CNRS, Inria, I3S, Université Côte d’Azur, France · School of Electrical and Computer Engineering, Tel Aviv University, Israel · Department of Informatics, King’s College London, UK +1
Optimizing Digital Therapeutic Interventions: Online Learning under Endogenous Adherence
May 22, 2026Eric Pulick, Stephanie Carpenter, Matthew Buman +1Dynamical Systems Multi-Armed Bandits
Department of Industrial and Systems Engineering, University of Wisconsin-Madison · College of Health Solutions, Arizona State University
Learning dynamical systems with biochemically informed neural ordinary differential equations
May 22, 2026Luis L. Fonseca, Reinhard C. Laubenbacher, Lucas BöttcherDynamical Systems Neural Processes
Laboratory for Systems Medicine, Department of Medicine, University of Florida, Gainesville, FL, USA · Department of Computational Science and Philosophy, Frankfurt School of Finance and Management, Frankfurt am Main, 60322, Germany
Detecting Metastable Basins in High Dimensions via Marginal Trajectory Distribution Discrimination
May 22, 2026Taj Jones-McCormickDynamical Systems Two-Sample Testing
Learning Dynamic Stability Landscapes in Synchronization Networks
May 22, 2026Christian Nauck, Junyou Zhu, Michael Lindner +1Dynamical Systems
Department of Complexity Science, Postdam Institute for Climate Impact Research, Potsdam, Germany · Machine Learning Group, Technical University of Berlin, 10587 Berlin, Germany · Department of Digital Transformation in Energy Systems, Institute of Energy Technology, Technical University of Berlin, Germany.
Optimization of randomized neural networks for transfer operator approximation
May 22, 2026Mohammad Tabish, Stefan KlusDynamical Systems Neural Network Optimization
Maxwell Institute for Mathematical Sciences, University of Edinburgh and Heriot–Watt University, Edinburgh, UK · School of Mathematical & Computer Sciences, Heriot–Watt University, Edinburgh, UK
Learning partially observed systems with neural Hamiltonian ordinary differential equations
May 22, 2026Sunniva Meltzer, Sølve Eidnes, Alexander Johannes StasikDynamical Systems Latent Dynamics Modeling
Department of Mathematics and Cybernetics, SINTEF Digital, Norway · Department of Physics, University of Oslo, Norway · Department of Data Science, Norwegian University of Life Science, Norway
Non-normal spectral signatures of instability in neural network training dynamics
May 22, 2026Souvik GhoshDynamical Systems Neural Network Optimization
Department of Physics, National Sun Yat-sen University, Kaohsiung 80424, Taiwan
Feature Lottery? A Bifurcation Theory of Concept Emergence
May 22, 2026Fuming YangDynamical Systems Representation Learning
MIT
Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning
May 20, 2026Benhao Huang, Zhengyang Geng, Zico KolterDynamical Systems Test-Time Scaling
CMU
Equilibrium Propagation and Hamiltonian Inference in the Diffusive Fitzhugh-Nagumo Model
May 20, 2026Jack KendallEquilibrium Propagation Dynamical Systems
Zyphra · San Francisco, CA
Safety-Critical Control for Smoothed Implicit Contact Dynamics
May 20, 2026Haegu Lee, Yitaek Kim, Christoffer SlothDynamical Systems Contact-Rich Robotic Manipulation
The Maersk Mc-Kinney Moller Institute, University of Southern Denmark, Denmark
Reconfigurable Nonlinear Photonic Networks for In-Situ Learning and Memory Formation via Driven-Dissipative Dynamics
May 19, 2026Isaac YorkeDynamical Systems Neural Network Memorization
Department of Engineering and Architecture, University of Parma, Parco Area delle Scienze 181/A I-43124 Parma, Italy
Convergence of Consensus-Based Particle Methods for Nonconvex Bi-Level Optimization
May 19, 2026Yutong Chao, Xudong Sun, Konstantin Riedl +2Dynamical Systems Zeroth-Order Optimization
Department of Computer Science Technical University of Munich · Munich Institute of Robotics and Machine Intelligence Technical University of Munich · Mathematical Institute University of Oxford
From Mean-Field Limits to Semiclassical Concentration: Global Convergence of the Canonical Evolutionary Strategy
May 19, 2026Matías Neto, Nicolás Garay, Luis Martí +1Dynamical Systems Evolutionary Optimization
Inria Chile Research Center, Av Apoquindo 2827, Las Condes, RM, Chile · Departamento de Ingeniería Matemática, Universidad de Chile, Santiago, Chile
Generative Adversarial Learning from Deterministic Processes
May 18, 2026Joris C. Kühl, Hanno GottschalkDynamical Systems Statistical Learning Theory
Institute of Mathematics, Technical University of Berlin, Germany

May 31, 2026, cs.LGJ/K move · Enter open · S save

DAGGER: Gradient-Free Construction of Transiently Amplifying Networks under Hard Connectivity Constraints

James C. Ferguson

The African Institute for Mathematical Sciences (Muizenberg, South Africa) · the Institute of Science and Technology Austria (Klosterneuburg, Austria)

Dynamical Systems Sparse Neural Networks

Open paper AskPDF arXiv

Many networks not only support but also rely on transient non-normal amplification, an orders-of-magnitude increase in the activity of an otherwise stable system. Constructing such networks under hard sign/sparsity/diagonal constraints -- the regime relevant for biological connectomes and structured RNN initializations -- has so far required either gradient-based local search with thousands of inner-loop eigendecompositions or Schur-form direct construction in an abstract basis that breaks the constraints under projection. Here we introduce DAGGER (Directed Acyclic Graph Guided Edge Reweighting), a gradient-free single-pass algorithm. Given a stable signed sparse matrix, DAGGER produces an output with the same sign, sparsity, and diagonal. A single scalar

β

controls a Wasserstein-2 budget that smoothly trades exact multiset preservation (

β= 0

) for amplification; peak amplification grows essentially without bound with

β

, empirically reaching

10^{10}

before numerical overflow. DAGGER matches or exceeds gradient-based methods at multiset preservation in a single forward pass -- 30-100

\times

fewer eigendecompositions than a typical gradient inner loop -- and at moderate

β

beats them by orders of magnitude with connectivity exactly preserved. We develop the algorithm, compare it to the existing methods and on a downstream signal-detection task, and examine the diagnostics that show why DAGGER is structurally different from other amplifying networks.

Momentum

Related topics

Latest papers 319