Self-organizing maps based on limit cycle attractors

Huang, Di-Wei; Gentili, Rodolphe J.; Reggia, James A.

doi:10.1016/j.neunet.2014.12.003

Cited by 15 publications

(18 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This claim appears to be consistent with at least some neurophysiological findings [40]. Of course, there are other types of attractor states in recurrent neural networks besides fixed activity patterns, such as periodically recurring activity patterns (limit cycles) [41], and they could also be potential computational correlates. In particular, it has been proposed that instantaneous activity states of a network are inadequate for realizing conscious experience-that qualia are correlated instead with a system's activity space trajectory [35].…”

Section: Past Suggestions For Computational Correlates Of Consciousnesssupporting

confidence: 88%

Exploring the Computational Explanatory Gap

2017

Self Cite

View full text Add to dashboard Cite

While substantial progress has been made in the field known as artificial consciousness, at the present time there is no generally accepted phenomenally conscious machine, nor even a clear route to how one might be produced should we decide to try. Here, we take the position that, from our computer science perspective, a major reason for this is a computational explanatory gap: our inability to understand/explain the implementation of high-level cognitive algorithms in terms of neurocomputational processing. We explain how addressing the computational explanatory gap can identify computational correlates of consciousness. We suggest that bridging this gap is not only critical to further progress in the area of machine consciousness, but would also inform the search for neurobiological correlates of consciousness and would, with high probability, contribute to demystifying the "hard problem" of understanding the mind-brain relationship. We compile a listing of previously proposed computational correlates of consciousness and, based on the results of recent computational modeling, suggest that the gating mechanisms associated with top-down cognitive control of working memory should be added to this list. We conclude that developing neurocognitive architectures that contribute to bridging the computational explanatory gap provides a credible and achievable roadmap to understanding the ultimate prospects for a conscious machine, and to a better understanding of the mind-brain problem in general.

show abstract

Section: Past Suggestions For Computational Correlates Of Consciousnesssupporting

confidence: 88%

Exploring the Computational Explanatory Gap

2017

Self Cite

View full text Add to dashboard Cite

show abstract

“…Notice that since each activity pattern a(t) depends on the last activity pattern a(t − 1), the activity pattern of a map changes with time and forms a dynamical system. Specifically, we have found that limit cycles are a prominent class of attractors that are learned via self-organization [8,9]. In order to generate steady joint command output, the oscillatory activity in the joint position map needs to be "smoothed out".…”

Section: Neural Architecturementioning

confidence: 99%

“…Although our previous work has provided some positive preliminary results, it is limited in that the architecture has not been used to generate outputs and the data size is limited (50 pairs of phoneme sequences and images) [9]. More importantly, the ability to generalize to new and unseen data, a critical indicator of a successful neurocognitive architecture, is not clear.…”

Section: Introductionmentioning

confidence: 99%

“…Moreover, each activity pattern typically contains only a single maximally activated node (i.e., a "winning" node), which is similar to a one-hot encoding scheme. Our previous work addresses this issue by proposing SOMs based on a limit cycle representation [8,9]. That is, instead of a fixed single-winner activity pattern, each input is encoded by a temporal sequence of multi-winner activity patterns that forms a limit cycle attractor.…”

Section: Introductionmentioning

confidence: 99%

“…The limit cycles are learned through self-organization rather than manually specified. We have recently shown that such a representation can be used to encode both static vectors [8] and temporal sequences [9]. A major distinction of our approach is that it takes into account the ongoing temporal transformation of a SOM's activity, rather than only a specific activity pattern occurring at a certain time.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Self-Organizing Map Architecture for Arm Reaching Based on Limit Cycle Attractors

Huang¹,

Gentili²,

Reggia³

2016

Proceedings of the 9th EAI International Conference on Bio-Inspired Information and Communications Technologies (Formerly BIONE

Self Cite

View full text Add to dashboard Cite

Creating and studying neurocognitive architectures is an active and increasing focus of research efforts. Based on our recent research that uses neural activity limit cycles in selforganizing maps (SOMs) to represent external stimuli, this study explores the use of such limit cycle attractors in a neurocognitive architecture for an open-loop arm reaching task. The goal is to learn to produce a static motor command for arm joints that moves the manipulator to a target spatial location, while the internal neural activity remains oscillatory. Unlike with static SOMs, stabilizing output based on changing neural activity becomes an important issue. Our architecture is also characterized by simple and forgiving timing requirements, meaning that the time of gating among neural components can be set relatively arbitrarily due to the repetitiveness of limit cycle activity. The results indicate that our architecture generalizes to unseen data, and that the overall performance is insensitive to exact gate timing.

show abstract