Artificial Intelligence


Supersymmetric Artificial Neural Network

Authors: Jordan Micah Bennett

Babies are great examples of some non-trivial basis for artificial general intelligence; babies are significant examples of biological baseis that are reasonably usable to inspire smart algorithms. The “Supersymmetric Artificial Neural Network” in deep learning (denoted φ(x, θ, θ)⊤w), espouses the importance of considering biological constraints in the aim of developing general machine learning models, pertinently, where babies' brains are observed to be pre-equipped with particular "physics priors", constituting specifically, the ability for babies to intuitively know laws of physics, while learning by reinforcement. It is palpable that the phrasing “intuitively know laws of physics” above, should not be confused for nobel laureate or physics undergrad aligned babies that for example, write or understand physics papers/exams; instead, the aforesaid phrasing simply conveys that babies' brains are pre-baked with ways to naturally exercise physics based expectations w.r.t. interactions with objects in their world, as indicated by Aimee Stahl and Lisa Feigenson. Outstandingly, the importance of recognizing underlying causal physics laws in learning models (although not via supermanifolds, as encoded in the “Supersymmetric Artificial Neural Network”), has recently been both demonstrated and separately echoed by Deepmind (See “Neuroscience-Inspired Artificial Intelligence“) and of late, distinctly emphasized by Yoshua Bengio (See the “Consciousness Prior”). Physics based object detectors like "Uetorch" use something called pooling to gain translation invariance over objects, so that the model learns regardless of where the object in the image is positioned, while instead, reinforcement models like "AtariQLearner" exclude pooling, because "AtariQLearner" requires translation variance, in order for Q learning to apply on the changing positions of the objects in pixels. Babies seem to be able to do both these activities. That said, an example of models that can deliver both translation invariance and variance at the same time, i.e. disentangled factors of variation, are called manifold learning frameworks (Bengio et al. ...). Given that cognitive science may be used to constrain machine learning models (similar to how firms like Deepmind often use cognitive science as a boundary on the deep learning models they produce) The " Supersymmetric Artificial Neural Network” is a uniquely disentanglable model that is constrained by cognitive science, in the direction of supermanifolds (See “Supersymmetric methods ... at brain scale”, Perez et al.), instead of state of the art manifold work by other authors. (Such as manifold work by Bengio et al., Lecun et al. or Michael Bronstein et al.) As such, the "Supersymmetric Artificial Neural Network" is yet another way to represent richer values in the weights of the model; because supersymmetric values can allow for more information to be captured about the input space. For example, supersymmetric systems can capture potential-partner signals, which are beyond the feature space of magnitude and phase signals learnt in typical real valued neural nets and deep complex neural networks respectively. Looking at the progression of ‘solution geometries’; going from SO(n) representation (such as Perceptron like models) to SU(n) representation (such as UnitaryRNNs) has guaranteed richer and richer representations in weight space of the artificial neural network, and hence better and better hypotheses were generatable. The Supersymmetric Artificial Neural Network explores a natural step forward, namely SU(m|n) representation. These supersymmetric biological brain representations (Perez et al.) can be represented by supercharge compatible special unitary notation SU(m|n), or φ(x, θ, `θ)Tw parameterized by θ, `θ, which are supersymmetric directions, unlike θ seen in the typical non-supersymmetric deep learning model. Notably, Supersymmetric values can encode or represent more information than the typical deep learning model, in terms of “partner potential” signals for example.

Comments: 12 Pages. Author Email: Author Website:

Download: PDF

Submission history

[v1] 2018-10-09 21:37:41

Unique-IP document downloads: 23 times is a pre-print repository rather than a journal. Articles hosted may not yet have been verified by peer-review and should be treated as preliminary. In particular, anything that appears to include financial or legal advice or proposed medical treatments should be treated with due caution. will not be responsible for any consequences of actions that result from any form of use of any documents on this website.

Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.

comments powered by Disqus