AI Paper Atlas

Research map / Time series / Causal and generative time series

Weight-Space Linear Recurrent Neural Networks

Roussel Desmond Nzoyem, Nawid Keshtmand, Enrique Crespo-Fernandez, Idriss TSAYEM, Raul Santos-Rodriguez, David A.W. Barton, Tom Deakin

ICLR 2026 · Time series › Causal and generative time series · approach: Dynamical system reconstruction · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add physical prior (mass-spring-damper form) in root networkWARP · MSDMSE: 0.94 -> 0.03Improved
add physical prior in root networkWARP · MSD-ZeroMSE: 0.32 -> 0.04Improved
embed sine formula with predicted phase in root networkWARP · SINE* (Small split)MSE: 2.77 -> 0.62Improved
use weight-space recurrence instead of GRUGRU baseline · LVMSE: 5.83 -> 4.72Improved
use weight-space recurrence instead of GRUGRU baseline · CelebA (L=300)BPD: 60.39 -> -0.043Improved
use weight-space recurrence instead of TransformerTransformer (TST) · SINE* (Small split)MSE: 1728 -> 2.77Improved
use WARP instead of graph-based STIDGCNSTIDGCN · PEMS08MAE: 13.45 -> 6.59Improved
use WARP instead of S4S4 · MNIST (L=600)BPD: 0.406 -> 0.416Hurt
use WARP instead of TransformerTransformer (TST) · MSDMSE: 0.34 -> 0.94Hurt
use WARP instead of LinOSSLinOSS · UEA EigenWormsaccuracy: 95.0 -> 70.93Hurt
use WARP instead of LRULRU · UEA Heartbeataccuracy: 78.1 -> 80.65Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy