Evolutionary Training and Abstraction Yields Algorithmic Generalization of Neural Computers

by   Daniel Tanneberg, et al.

A key feature of intelligent behaviour is the ability to learn abstract strategies that scale and transfer to unfamiliar problems. An abstract strategy solves every sample from a problem class, no matter its representation or complexity – like algorithms in computer science. Neural networks are powerful models for processing sensory data, discovering hidden patterns, and learning complex functions, but they struggle to learn such iterative, sequential or hierarchical algorithmic strategies. Extending neural networks with external memories has increased their capacities in learning such strategies, but they are still prone to data variations, struggle to learn scalable and transferable solutions, and require massive training data. We present the Neural Harvard Computer (NHC), a memory-augmented network based architecture, that employs abstraction by decoupling algorithmic operations from data manipulations, realized by splitting the information flow and separated modules. This abstraction mechanism and evolutionary training enable the learning of robust and scalable algorithmic solutions. On a diverse set of 11 algorithms with varying complexities, we show that the NHC reliably learns algorithmic solutions with strong generalization and abstraction: perfect generalization and scaling to arbitrary task configurations and complexities far beyond seen during training, and being independent of the data representation and the task domain.


page 1

page 2

page 3

page 6

page 7

page 8


Learning Algorithmic Solutions to Symbolic Planning Tasks with a Neural Computer

A key feature of intelligent behavior is the ability to learn abstract s...

Learning Continuous Chaotic Attractors with a Reservoir Computer

Neural systems are well known for their ability to learn and store infor...

DeepAbstract: Neural Network Abstraction for Accelerating Verification

While abstraction is a classic tool of verification to scale it up, it i...

A Memory-Augmented Neural Network Model of Abstract Rule Learning

Human intelligence is characterized by a remarkable ability to infer abs...

Emergent Symbols through Binding in External Memory

A key aspect of human intelligence is the ability to infer abstract rule...

Improving the Universality and Learnability of Neural Programmer-Interpreters with Combinator Abstraction

To overcome the limitations of Neural Programmer-Interpreters (NPI) in i...

A Group-Theoretic Approach to Abstraction: Hierarchical, Interpretable, and Task-Free Clustering

Abstraction plays a key role in concept learning and knowledge discovery...

Please sign up or login with your details

Forgot password? Click here to reset