Markov transitions between attractor states in a recurrent Neural network

Jeremy Bernstein; Ishita Dasgupta; David Rolnick; Haim Sompolinsky

Markov transitions between attractor states in a recurrent Neural network

Jeremy Bernstein, Ishita Dasgupta, David Rolnick, Haim Sompolinsky

Racah Institute of Physics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Scopus citations

Abstract

Stochasticity is an essential part of explaining the world. Increasingly, neuroscientists and cognitive scientists are identifying mechanisms whereby the brain uses probabilistic reasoning in representational, predictive, and generative settings. But stochasticity is not always useful: robust perception and memory retrieval require representations that are immune to corruption by stochastic noise. In an effort to combine these robust representations with stochastic computation, we present an architecture that generalizes traditional recurrent attractor networks to follow probabilistic Markov dynamics between stable and noise-resistant fixed points.

Original language	English
Title of host publication	SS-17-01
Subtitle of host publication	Artificial Intelligene for the Social Good; SS-17-02: Computational Construction Grammar and Natural Language Understanding; SS-17-03: Computational Context: Why It's Important, What It Means, and Can It Be Computed?; SS-17-04: Designing the User Experience of Machine Learning Systems; SS-17-05: Interactive Multisensory Object Perception for Embodied Agents; SS-17-06: Learning from Observation of Humans; SS-17-07: Science of Intelligence: Computational Principles of Natural and Artificial Intelligence; SS-17-08: Wellbeing AI: From Machine Learning to Subjectivity Oriented Computing
Publisher	AI Access Foundation
Pages	529-533
Number of pages	5
ISBN (Electronic)	9781577357797
State	Published - 2017
Event	2017 AAAI Spring Symposium - Stanford, United States Duration: 27 Mar 2017 → 29 Mar 2017

Publication series

Name	AAAI Spring Symposium - Technical Report
Volume	SS-17-01 - SS-17-08

Conference

Conference	2017 AAAI Spring Symposium
Country/Territory	United States
City	Stanford
Period	27/03/17 → 29/03/17

Bibliographical note

Publisher Copyright:
© Copyright 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.

Cite this

Bernstein, J., Dasgupta, I., Rolnick, D., & Sompolinsky, H. (2017). Markov transitions between attractor states in a recurrent Neural network. In SS-17-01: Artificial Intelligene for the Social Good; SS-17-02: Computational Construction Grammar and Natural Language Understanding; SS-17-03: Computational Context: Why It's Important, What It Means, and Can It Be Computed?; SS-17-04: Designing the User Experience of Machine Learning Systems; SS-17-05: Interactive Multisensory Object Perception for Embodied Agents; SS-17-06: Learning from Observation of Humans; SS-17-07: Science of Intelligence: Computational Principles of Natural and Artificial Intelligence; SS-17-08: Wellbeing AI: From Machine Learning to Subjectivity Oriented Computing (pp. 529-533). (AAAI Spring Symposium - Technical Report; Vol. SS-17-01 - SS-17-08). AI Access Foundation.

Bernstein, Jeremy ; Dasgupta, Ishita ; Rolnick, David et al. / Markov transitions between attractor states in a recurrent Neural network. SS-17-01: Artificial Intelligene for the Social Good; SS-17-02: Computational Construction Grammar and Natural Language Understanding; SS-17-03: Computational Context: Why It's Important, What It Means, and Can It Be Computed?; SS-17-04: Designing the User Experience of Machine Learning Systems; SS-17-05: Interactive Multisensory Object Perception for Embodied Agents; SS-17-06: Learning from Observation of Humans; SS-17-07: Science of Intelligence: Computational Principles of Natural and Artificial Intelligence; SS-17-08: Wellbeing AI: From Machine Learning to Subjectivity Oriented Computing. AI Access Foundation, 2017. pp. 529-533 (AAAI Spring Symposium - Technical Report).

@inproceedings{5ea128856329469a8e4e5a556fb52e4b,

title = "Markov transitions between attractor states in a recurrent Neural network",

abstract = "Stochasticity is an essential part of explaining the world. Increasingly, neuroscientists and cognitive scientists are identifying mechanisms whereby the brain uses probabilistic reasoning in representational, predictive, and generative settings. But stochasticity is not always useful: robust perception and memory retrieval require representations that are immune to corruption by stochastic noise. In an effort to combine these robust representations with stochastic computation, we present an architecture that generalizes traditional recurrent attractor networks to follow probabilistic Markov dynamics between stable and noise-resistant fixed points.",

author = "Jeremy Bernstein and Ishita Dasgupta and David Rolnick and Haim Sompolinsky",

note = "Publisher Copyright: {\textcopyright} Copyright 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.; 2017 AAAI Spring Symposium ; Conference date: 27-03-2017 Through 29-03-2017",

year = "2017",

language = "אנגלית",

series = "AAAI Spring Symposium - Technical Report",

publisher = "AI Access Foundation",

pages = "529--533",

booktitle = "SS-17-01",

address = "ארצות הברית",

}

Bernstein, J, Dasgupta, I, Rolnick, D & Sompolinsky, H 2017, Markov transitions between attractor states in a recurrent Neural network. in SS-17-01: Artificial Intelligene for the Social Good; SS-17-02: Computational Construction Grammar and Natural Language Understanding; SS-17-03: Computational Context: Why It's Important, What It Means, and Can It Be Computed?; SS-17-04: Designing the User Experience of Machine Learning Systems; SS-17-05: Interactive Multisensory Object Perception for Embodied Agents; SS-17-06: Learning from Observation of Humans; SS-17-07: Science of Intelligence: Computational Principles of Natural and Artificial Intelligence; SS-17-08: Wellbeing AI: From Machine Learning to Subjectivity Oriented Computing. AAAI Spring Symposium - Technical Report, vol. SS-17-01 - SS-17-08, AI Access Foundation, pp. 529-533, 2017 AAAI Spring Symposium, Stanford, United States, 27/03/17.

Markov transitions between attractor states in a recurrent Neural network. / Bernstein, Jeremy; Dasgupta, Ishita; Rolnick, David et al.
SS-17-01: Artificial Intelligene for the Social Good; SS-17-02: Computational Construction Grammar and Natural Language Understanding; SS-17-03: Computational Context: Why It's Important, What It Means, and Can It Be Computed?; SS-17-04: Designing the User Experience of Machine Learning Systems; SS-17-05: Interactive Multisensory Object Perception for Embodied Agents; SS-17-06: Learning from Observation of Humans; SS-17-07: Science of Intelligence: Computational Principles of Natural and Artificial Intelligence; SS-17-08: Wellbeing AI: From Machine Learning to Subjectivity Oriented Computing. AI Access Foundation, 2017. p. 529-533 (AAAI Spring Symposium - Technical Report; Vol. SS-17-01 - SS-17-08).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Markov transitions between attractor states in a recurrent Neural network

AU - Bernstein, Jeremy

AU - Dasgupta, Ishita

AU - Rolnick, David

AU - Sompolinsky, Haim

PY - 2017

Y1 - 2017

N2 - Stochasticity is an essential part of explaining the world. Increasingly, neuroscientists and cognitive scientists are identifying mechanisms whereby the brain uses probabilistic reasoning in representational, predictive, and generative settings. But stochasticity is not always useful: robust perception and memory retrieval require representations that are immune to corruption by stochastic noise. In an effort to combine these robust representations with stochastic computation, we present an architecture that generalizes traditional recurrent attractor networks to follow probabilistic Markov dynamics between stable and noise-resistant fixed points.

AB - Stochasticity is an essential part of explaining the world. Increasingly, neuroscientists and cognitive scientists are identifying mechanisms whereby the brain uses probabilistic reasoning in representational, predictive, and generative settings. But stochasticity is not always useful: robust perception and memory retrieval require representations that are immune to corruption by stochastic noise. In an effort to combine these robust representations with stochastic computation, we present an architecture that generalizes traditional recurrent attractor networks to follow probabilistic Markov dynamics between stable and noise-resistant fixed points.

UR - http://www.scopus.com/inward/record.url?scp=85028723568&partnerID=8YFLogxK

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:85028723568

T3 - AAAI Spring Symposium - Technical Report

SP - 529

EP - 533

BT - SS-17-01

PB - AI Access Foundation

T2 - 2017 AAAI Spring Symposium

Y2 - 27 March 2017 through 29 March 2017

ER -

Bernstein J, Dasgupta I, Rolnick D, Sompolinsky H. Markov transitions between attractor states in a recurrent Neural network. In SS-17-01: Artificial Intelligene for the Social Good; SS-17-02: Computational Construction Grammar and Natural Language Understanding; SS-17-03: Computational Context: Why It's Important, What It Means, and Can It Be Computed?; SS-17-04: Designing the User Experience of Machine Learning Systems; SS-17-05: Interactive Multisensory Object Perception for Embodied Agents; SS-17-06: Learning from Observation of Humans; SS-17-07: Science of Intelligence: Computational Principles of Natural and Artificial Intelligence; SS-17-08: Wellbeing AI: From Machine Learning to Subjectivity Oriented Computing. AI Access Foundation. 2017. p. 529-533. (AAAI Spring Symposium - Technical Report).

Markov transitions between attractor states in a recurrent Neural network

Abstract

Publication series

Conference

Bibliographical note

Other files and links

Fingerprint

Cite this