Peer-reviewedLicence not statedNeural Computation
Shaping Slow Manifolds in Equivariant Recurrent Neural Networks
4 affiliations
- 1Laboratoire de Neurosciences Cognitives et Computationnelles, INSERM U960, École Normale Supérieure–PSL Research University, 75005 Paris, France [email protected]
- 2Laboratoire de Neurosciences Cognitives et Computationnelles, INSERM U960, École Normale Supérieure–PSL Research University, 75005 Paris, France [email protected]
- 3SISSA—Theoretical and Scientific Data Science, 34136 Trieste, Italy [email protected]
- 4Laboratoire de Neurosciences Cognitives et Computationnelles, INSERM U960, École Normale Supérieure–PSL Research University, 75005 Paris, France
Related
- Petimot: A Novel Framework for Inferring Protein Motions from Sparse Data Using SE(3)-Equivariant Graph Neural Networks
- CohAlign: Predicting the first-order noise susceptibility of gradients in equivariant quantum neural networks
- Operator-valued free probability applied to recurrent neural networks with heterogeneity in local and global circuit structures
- High-rank connectivity scaffolds support precision and generalisation in recurrent neural networks
- Dynamics of random recurrent neural networks: colored noise and sublinear activations
BibTeX
@article{bernardo2026shaping,
title = {{Shaping Slow Manifolds in Equivariant Recurrent Neural Networks}},
author = {Arianna Di Bernardo and Adrian Valente and Francesca Mastrogiuseppe and Srdjan Ostojic},
journal = {Neural Computation},
year = {2026},
month = sep,
doi = {10.1162/neco.a.1577},
url = {http://arxiv.org/abs/2511.04802},
}
RIS
TY - JOUR
TI - Shaping Slow Manifolds in Equivariant Recurrent Neural Networks
AU - Arianna Di Bernardo
AU - Adrian Valente
AU - Francesca Mastrogiuseppe
AU - Srdjan Ostojic
PY - 2026
DA - 2026/09/22
JO - Neural Computation
DO - 10.1162/neco.a.1577
UR - http://arxiv.org/abs/2511.04802
SN - 0899-7667
SN - 1530-888X
ER -