Connectionists: New paper on why modules evolve, and how to evolve modular neural networks

[Tsvi Achler]

I would like to bring to your attention another way to implement
modularity separate from sparseness.

Modularity as defined here is the ability to add or modify a weight or
a neuron in a network neural network without potentially changing the
integrity of representation of other neurons in the neural network.

Restating this in mathematical terms: modularity is the ability to
modify a single fixed point by modifying a single neuron or weight,
without changing the other fixed points.

In “feedforward” neural networks a single change in weight can change
multiple fixed points.  One way to avoid this is to make the network
sparse, evolve more-modular networks.  A completely different way to
implement modularity is to perform dynamics during recognition.

Such networks use reentrant “feedforward-feedback” connections during
recognition.  In other words, it is common in neural networks to
iterate to find the best weights during learning, and use feedforward
weights during recognition.  Instead for better modularity, a similar
iterating mechanism is implemented during recognition to find
activation (not modify weights).  The dynamic mechanism can be
implemented using symmetric reentrant inhibitory connections.  Once
this is achieved, if single neuron or a reentrant weight pair is
modified, it only changes a single fixed point.

This may be a bit confusing at first especially since not many neural
network models truly use reentrant dynamics during recognition.
Dynamics are commonly used during learning and to determine
sparseness.  Please feel free to ask me questions.  I would be happy
to give a talk about this and see references below.

Achler, T., Supervised Generative Reconstruction: An Efficient Way To
Flexibly Store and Recognize Patterns, Arxiv 2012,
http://arxiv.org/pdf/1112.2988v2

Achler, T., Non-Oscillatory Dynamics to Disambiguate Pattern Mixtures,
Chapter 4 in Relevance of the Time Domain to Neural Network Models
2011, http://reason.cs.uiuc.edu/tsvi/TimeDomain_Chapter.pdf

Achler, T.,Towards Bridging the Gap between Pattern Recognition and
Symbolic Representations Within Neural Networks, Neural-Symbolic
Learning and Reasoning, AAAI-2012,
http://reason.cs.uiuc.edu/tsvi/nesy2012.pdf

Sincerely,

Tsvi Achler MD/PhD



On Sat, Feb 9, 2013 at 6:14 PM, Jeff Clune <jeffclune at uwyo.edu> wrote:
> Hello all,
>
> I believe that many in the neuroscience community will be interested in a new paper that sheds light on why modularity evolves in biological networks, including neural networks. The same discovery also provides AI researchers a simple technique for evolving neural networks that are modular and have increased evolvability, meaning that they adapt faster to new environments.
>
> Cite: Clune J, Mouret J-B, Lipson H (2013) The evolutionary origins of modularity. Proceedings of the Royal Society B. 280: 20122863. http://dx.doi.org/10.1098/rspb.2012.2863 (pdf)
>
> Abstract: A central biological question is how natural organisms are so evolvable (capable of quickly adapting to new environments). A key driver of evolvability is the widespread modularity of biological networks—their organization as functional, sparsely connected subunits—but there is no consensus regarding why modularity itself evolved. Although most hypotheses assume indirect selection for evolvability, here we demonstrate that the ubiquitous, direct selection pressure to reduce the cost of connections between network nodes causes the emergence of modular networks. Computational evolution experiments with selection pressures to maximize network performance and minimize connection costs yield networks that are significantly more modular and more evolvable than control experiments that only select for performance. These results will catalyse research in numerous disciplines, such as neuroscience and genetics, and enhance our ability to harness evolution for engineering pu!
>  rposes.
>
> Video: http://www.youtube.com/watch?feature=player_embedded&v=SG4_aW8LMng
>
> There has been some nice coverage of this work in the popular press, in case you are interested:
>
> • National Geographic: http://phenomena.nationalgeographic.com/2013/01/30/the-parts-of-life/
> • MIT's Technology Review: http://www.technologyreview.com/view/428504/computer-scientists-reproduce-the-evolution-of-evolvability/
> • Fast Company: http://www.fastcompany.com/3005313/evolved-brains-robots-creep-closer-animal-learning
> • Cornell Chronicle: http://www.news.cornell.edu/stories/Jan13/modNetwork.html
> • ScienceDaily: http://www.sciencedaily.com/releases/2013/01/130130082300.htm
>
> I hope you enjoy the work. Please let me know if you have any questions.
>
> Best regards,
> Jeff Clune
>
> Assistant Professor
> Computer Science
> University of Wyoming
> jeffclune at uwyo.edu
> jeffclune.com
>
>


On Sat, Feb 9, 2013 at 6:14 PM, Jeff Clune <jeffclune at uwyo.edu> wrote:
> Hello all,
>
> I believe that many in the neuroscience community will be interested in a new paper that sheds light on why modularity evolves in biological networks, including neural networks. The same discovery also provides AI researchers a simple technique for evolving neural networks that are modular and have increased evolvability, meaning that they adapt faster to new environments.
>
> Cite: Clune J, Mouret J-B, Lipson H (2013) The evolutionary origins of modularity. Proceedings of the Royal Society B. 280: 20122863. http://dx.doi.org/10.1098/rspb.2012.2863 (pdf)
>
> Abstract: A central biological question is how natural organisms are so evolvable (capable of quickly adapting to new environments). A key driver of evolvability is the widespread modularity of biological networks—their organization as functional, sparsely connected subunits—but there is no consensus regarding why modularity itself evolved. Although most hypotheses assume indirect selection for evolvability, here we demonstrate that the ubiquitous, direct selection pressure to reduce the cost of connections between network nodes causes the emergence of modular networks. Computational evolution experiments with selection pressures to maximize network performance and minimize connection costs yield networks that are significantly more modular and more evolvable than control experiments that only select for performance. These results will catalyse research in numerous disciplines, such as neuroscience and genetics, and enhance our ability to harness evolution for engineering pu!
>  rposes.
>
> Video: http://www.youtube.com/watch?feature=player_embedded&v=SG4_aW8LMng
>
> There has been some nice coverage of this work in the popular press, in case you are interested:
>
> • National Geographic: http://phenomena.nationalgeographic.com/2013/01/30/the-parts-of-life/
> • MIT's Technology Review: http://www.technologyreview.com/view/428504/computer-scientists-reproduce-the-evolution-of-evolvability/
> • Fast Company: http://www.fastcompany.com/3005313/evolved-brains-robots-creep-closer-animal-learning
> • Cornell Chronicle: http://www.news.cornell.edu/stories/Jan13/modNetwork.html
> • ScienceDaily: http://www.sciencedaily.com/releases/2013/01/130130082300.htm
>
> I hope you enjoy the work. Please let me know if you have any questions.
>
> Best regards,
> Jeff Clune
>
> Assistant Professor
> Computer Science
> University of Wyoming
> jeffclune at uwyo.edu
> jeffclune.com
>
>

On Sat, Feb 9, 2013 at 6:14 PM, Jeff Clune <jeffclune at uwyo.edu> wrote:
> Hello all,
>
> I believe that many in the neuroscience community will be interested in a new paper that sheds light on why modularity evolves in biological networks, including neural networks. The same discovery also provides AI researchers a simple technique for evolving neural networks that are modular and have increased evolvability, meaning that they adapt faster to new environments.
>
> Cite: Clune J, Mouret J-B, Lipson H (2013) The evolutionary origins of modularity. Proceedings of the Royal Society B. 280: 20122863. http://dx.doi.org/10.1098/rspb.2012.2863 (pdf)
>
> Abstract: A central biological question is how natural organisms are so evolvable (capable of quickly adapting to new environments). A key driver of evolvability is the widespread modularity of biological networks—their organization as functional, sparsely connected subunits—but there is no consensus regarding why modularity itself evolved. Although most hypotheses assume indirect selection for evolvability, here we demonstrate that the ubiquitous, direct selection pressure to reduce the cost of connections between network nodes causes the emergence of modular networks. Computational evolution experiments with selection pressures to maximize network performance and minimize connection costs yield networks that are significantly more modular and more evolvable than control experiments that only select for performance. These results will catalyse research in numerous disciplines, such as neuroscience and genetics, and enhance our ability to harness evolution for engineering pu!
>  rposes.
>
> Video: http://www.youtube.com/watch?feature=player_embedded&v=SG4_aW8LMng
>
> There has been some nice coverage of this work in the popular press, in case you are interested:
>
> • National Geographic: http://phenomena.nationalgeographic.com/2013/01/30/the-parts-of-life/
> • MIT's Technology Review: http://www.technologyreview.com/view/428504/computer-scientists-reproduce-the-evolution-of-evolvability/
> • Fast Company: http://www.fastcompany.com/3005313/evolved-brains-robots-creep-closer-animal-learning
> • Cornell Chronicle: http://www.news.cornell.edu/stories/Jan13/modNetwork.html
> • ScienceDaily: http://www.sciencedaily.com/releases/2013/01/130130082300.htm
>
> I hope you enjoy the work. Please let me know if you have any questions.
>
> Best regards,
> Jeff Clune
>
> Assistant Professor
> Computer Science
> University of Wyoming
> jeffclune at uwyo.edu
> jeffclune.com
>
>