Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Axiomatic scalable neurocontroller analysis via the Shapley value.

Alon Keinan1, Ben Sandbank, Claus C Hilgetag

  • 1School of Computer Science, Tel-Aviv University, Tel-Aviv, Israel. keinanak@post.tau.ac.il

Artificial Life
|July 25, 2006
PubMed
Summary

Researchers developed scalable multi-perturbation Shapley value analysis (MSA) to understand neural mechanisms in autonomous agents. This method efficiently deciphers causal function localization in complex networks, advancing artificial intelligence research.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Controlling reciprocity in binary and weighted networks: A novel density-conserving approach.

Chaos (Woodbury, N.Y.)·2026
Same author

Simple rules for simple models.

Physics of life reviews·2025
Same author

Pathway-Specific Ultrastructure of Thalamocortical Synapses in Mouse Somatosensory Area S2.

The Journal of neuroscience : the official journal of the Society for Neuroscience·2025
Same author

The coming decade of digital brain research: A vision for neuroscience at the intersection of technology and computing.

Imaging neuroscience (Cambridge, Mass.)·2025
Same author

Ultrastructural sublaminar-specific diversity of excitatory synaptic boutons in layer 1 of the adult human temporal lobe neocortex.

eLife·2025
Same author

Analysis of clinical anatomical correlates of motor deficits in stroke by multivariate lesion inference based on game theory.

Frontiers in neuroscience·2025

Area of Science:

  • Neuroscience
  • Artificial Intelligence
  • Game Theory

Background:

  • Deciphering neural mechanisms is crucial for advancing neurally driven evolved autonomous agents.
  • Existing methods for causal function localization are limited in scalability.
  • Understanding network element contributions and interactions is key to explaining agent behavior.

Purpose of the Study:

  • To develop scalable variants of the multi-perturbation Shapley value analysis (MSA).
  • To enable efficient causal function localization in large, complex neural networks.
  • To improve the analysis of neurally driven autonomous agents.

Main Methods:

  • Developed novel scalable variants of the multi-perturbation Shapley value analysis (MSA).
  • Applied game theory concepts for axiomatic and rigorous deduction of causal function localization.

Related Experiment Videos

  • Utilized multiple-perturbation data for network analysis.

    • Main Results:

    • Introduced scalable MSA variants capable of analyzing large complex networks efficiently.
    • Demonstrated successful application of MSA and its variants on neurocontrollers.
    • Analyzed neurocontrollers with up to 100 neural elements performing a food foraging task.

    Conclusions:

    • Scalable MSA variants overcome previous limitations in analyzing large-scale neurocontrollers.
    • The MSA framework provides a rigorous method for understanding neural mechanisms in complex systems.
    • This work advances the field of neurally driven evolved autonomous agents by enabling detailed behavioral analysis.