Work

Dynamical Similarity in Multitasking RNNs: Identifying Shared Motifs Across Task Periods

Year: 2025

Type: article

Abstract: Biological networks exhibit remarkable flexibility, enabling efficient multitasking. This study investigates the role of dynamical motifs in facilitating this adaptability within recurrent neural networks (RNNs). Instead of focusing on geometric structures, we employ Dynamic Similarity Analysis (DSA) to reveal underlying dynamic patterns. applying DSA to the internal states of RNNs across 15 tasks, we identify the delayed response motif as particularly distinct, characterized by unique rotational properties. Our findings highlight DSA’s ability to uncover differences overlooked by geometric methods, offering deeper insights into the computational mechanisms underlying both biological and artificial intelligence systems.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.15126464

Authors: Muhammad, Rabia, Yamauti, Seidi Yonamine, Gupta, Akanksha, Khajehnejad, Moein

Institutions: Neuromatch, Universidade Federal de São Paulo, Inserm, Aix-Marseille Université, Institut de Neurosciences des Systèmes, Monash University

Language: English

Topic: Neural Networks and Applications

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

OpenAlex ID: W6949575682

FWCI:

Cites:

Cited by:

Related to: