Object Files and Schemata: Factorizing Declarative and Procedural Knowledge in Dynamical Systems

Abstract

Modeling a structured, dynamic environment like a video game requires keeping track of the objects and their states declarative knowledge) as well as predicting how objects behave (procedural knowledge). Black-box models with a monolithic hidden state often fail to apply procedural knowledge consistently and uniformly, i.e., they lack systematicity. For example, in a video game, correct prediction of one enemy's trajectory does not ensure correct prediction of another's. We address this issue via an architecture that factorizes declarative and procedural knowledge and that imposes modularity within each form of knowledge. The architecture consists of active modules called object files that maintain the state of a single object and invoke passive external knowledge sources called schemata that prescribe state updates. To use a video game as an illustration, two enemies of the same type will share schemata but will have separate object files to encode their distinct state (e.g., health, position). We propose to use attention to determine which object files to update, the selection of schemata, and the propagation of information between object files. The resulting architecture is a drop-in replacement conforming to the same input-output interface as normal recurrent networks (e.g., LSTM, GRU) yet achieves substantially better generalization on environments that have multiple object tokens of the same type, including a challenging intuitive physics benchmark.

arXiv commentType/Token Distinction in Deep learning Framework

References in corpus (12)

• Adam: A Method for Stochastic Optimization
• Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
• Categorical Reparameterization with Gumbel-Softmax
• End-To-End Memory Networks
• PathNet: Evolution Channels Gradient Descent in Super Neural Networks
• Neural Programmer-Interpreters
• Relational Neural Expectation Maximization: Unsupervised Discovery of Objects and their Interactions
• Recurrent Independent Mechanisms
• Routing Networks and the Challenges of Modular and Compositional Computation
• Routing Networks: Adaptive Selection of Non-linear Functions for Multi-Task Learning
• Modular Networks: Learning to Decompose Neural Computation
• Relational recurrent neural networks

Cited by in corpus (8)

• Coordination Among Neural Modules Through a Shared Global Workspace
• Discrete-Valued Neural Communication
• Are Neural Nets Modular? Inspecting Functional Modularity Through Differentiable Weight Masks
• Dynamic Inference with Neural Interpreters
• Transformers with Competitive Ensembles of Independent Mechanisms
• Systematic Evaluation of Causal Discovery in Visual Model Based Reinforcement Learning
• Compositional Attention: Disentangling Search and Retrieval
• Language-Mediated, Object-Centric Representation Learning