This paper presents EINCASM, a prototype system employing a novel framework for studying emergent intelligence in organisms resembling slime molds. EINCASM evolves neural cellular automata with NEAT to maximize cell growth constrained by nutrient and energy costs. These organisms capitalize physically simulated fluid to transport nutrients and chemical-like signals to orchestrate growth and adaptation to complex, changing environments. Our framework builds the foundation for studying how the presence of puzzles, physics, communication, competition and dynamic open-ended environments contribute to the emergence of intelligent behavior. We propose preliminary tests for intelligence in such organisms and suggest future work for more powerful systems employing EINCASM to better understand intelligence in distributed dynamical systems.

Neural Cellular Automata (NCA) models have shown remarkable capacity for pattern formation and complex global behaviors stemming from local coordination. However, in the original implementation of NCA, cells are incapable of adjusting their own orientation, and it is the responsibility of the model designer to orient them externally. A recent isotropic variant of NCA (Growing Isotropic Neural Cellular Automata) makes the model orientation-independent - cells can no longer tell up from down, nor left from right - by removing its dependency on perceiving the gradient of spatial states in its neighborhood. In this work, we revisit NCA with a different approach: we make each cell responsible for its own orientation by allowing it to "turn" as determined by an adjustable internal state. The resulting Steerable NCA contains cells of varying orientation embedded in the same pattern. We observe how, while Isotropic NCA are orientation-agnostic, Steerable NCA have chirality: they have a predetermined left-right symmetry. We therefore show that we can train Steerable NCA in similar but simpler ways than their Isotropic variant by: (1) breaking symmetries using only two seeds, or (2) introducing a rotation-invariant training objective and relying on asynchronous cell updates to break the up-down symmetry of the system.

Federated machine learning is growing fast in academia and industries as a solution to solve data hungriness and privacy issues in machine learning. Being a widely distributed system, federated machine learning requires various system design thinking. To better design a federated machine learning system, researchers have introduced multiple patterns and tactics that cover various system design aspects. However, the multitude of patterns leaves the designers confused about when and which pattern to adopt. In this paper, we present a set of decision models for the selection of patterns for federated machine learning architecture design based on a systematic literature review on federated machine learning, to assist designers and architects who have limited knowledge of federated machine learning. Each decision model maps functional and non-functional requirements of federated machine learning systems to a set of patterns. We also clarify the drawbacks of the patterns. We evaluated the decision models by mapping the decision patterns to concrete federated machine learning architectures by big tech firms to assess the models' correctness and usefulness. The evaluation results indicate that the proposed decision models are able to bring structure to the federated machine learning architecture design process and help explicitly articulate the design rationale.

In this paper, we propose shot optimization method for QML models at the expense of minimal impact on model performance. We use classification task as a test case for MNIST and FMNIST datasets using a hybrid quantum-classical QML model. First, we sweep the number of shots for short and full versions of the dataset. We observe that training the full version provides 5-6% higher testing accuracy than short version of dataset with up to 10X higher number of shots for training. Therefore, one can reduce the dataset size to accelerate the training time. Next, we propose adaptive shot allocation on short version dataset to optimize the number of shots over training epochs and evaluate the impact on classification accuracy. We use a (a) linear function where the number of shots reduce linearly with epochs, and (b) step function where the number of shots reduce in step with epochs. We note around 0.01 increase in loss and around 4% (1%) reduction in testing accuracy for reduction in shots by up to 100X (10X) for linear (step) shot function compared to conventional constant shot function for MNIST dataset, and 0.05 increase in loss and around 5-7% (5-7%) reduction in testing accuracy with similar reduction in shots using linear (step) shot function on FMNIST dataset. For comparison, we also use the proposed shot optimization methods to perform ground state energy estimation of different molecules and observe that step function gives the best and most stable ground state energy prediction at 1000X less number of shots.

Consider a team of three guards (in green) trying to defend a fort from a team of three attackers (in red) in Figure 1. In this "Fort Attack" (FA) domain, each agent can move in one of four cardinal directions with a particular velocity, rotate clockwise or anticlockwise, shoot at an opponent within a given range, or do nothing. Each agent may have partial or full knowledge of the state of the world (e.g., location, status of each agent) at each step, but it has no prior experience of working with the other agents. Also, each agent may have limited (or no) ability to communicate with others. An episode ends when all members of a team are eliminated, an attacker reaches the fort, or the guards protect the fort for a sufficient time period.

Previous study of cellular automata and random Boolean networks has shown emergent behavior occurring at the edge of chaos where the randomness (disorder) of internal connections is set to an intermediate critical value. The value at which maximal emergent behavior occurs has been observed to be inversely related to the total number of interconnected elements, the neighborhood size. However, different equations predict different values. This paper presents a study of one-dimensional cellular automata (1DCA) verifying the general relationship but finding a more precise correlation with the radius of the neighborhood rather than neighborhood size. Furthermore, the critical value of the emergent regime is observed to be very close to 1/e hinting at the discovery of a fundamental characteristic of emergent systems.

The SIMD parallelism of the Connection Machine (eM) allows the construction of neural network simulations by the use of simple data and control structures. Two approaches are described which allow parallel computation of a model's nonlinear functions, parallel modification of a model's weights, and parallel propagation of a model's activation and error. Each approach also allows a model's interconnect structure to be physically dynamic. A Hopfield model is implemented with each approach at six sizes over the same number of CM processors to provide a performance comparison.

We describe a multi-network, or modular, connectionist architecture that captures that fact that many tasks have structure at a level of granularity intermediate to that assumed by local and global function approximation schemes. The main innovation of the architecture is that it combines associative and competitive learning in order to learn task decompositions. A task decomposition is discovered by forcing the networks comprising the architecture to compete to learn the training patterns. As a result of the competition, different networks learn different training patterns and, thus, learn to partition the input space. The performance of the architecture on a "what" and "where" vision task and on a multi-payload robotics task are presented.

We have trained networks of E - II units with short-range connec(cid:173) tions to simulate simple cellular automata that exhibit complex or chaotic behaviour. Three levels of learning are possible (in decreas(cid:173) ing order of difficulty): learning the underlying automaton rule, learning asymptotic dynamical behaviour, and learning to extrap(cid:173) olate the training history. The levels of learning achieved with and without weight sharing for different automata provide new insight into their dynamics.

We study the synthesis of neural coding, selective attention and percep- tual decision making. A hierarchical neural architecture is proposed, which implements Bayesian integration of noisy sensory input and top- down attentional priors, leading to sound perceptual discrimination. The model offers an explicit explanation for the experimentally observed modulation that prior information in one stimulus feature (location) can have on an independent feature (orientation). The network's intermediate levels of representation instantiate known physiological properties of vi- sual cortical neurons. The model also illustrates a possible reconciliation of cortical and neuromodulatory representations of uncertainty.