Neural Algorithmic Reasoners are Implicit Planners

Abstract

Implicit planning has emerged as an elegant technique for combining learned models of the world with end-to-end model-free reinforcement learning. We study the class of implicit planners inspired by value iteration, an algorithm that is guaranteed to yield perfect policies in fully-specified tabular environments. We find that prior approaches either assume that the environment is provided in such a tabular form -- which is highly restrictive -- or infer "local neighbourhoods" of states to run value iteration over -- for which we discover an algorithmic bottleneck effect. This effect is caused by explicitly running the planning algorithm based on scalar predictions in every state, which can be harmful to data efficiency if such scalars are improperly predicted. We propose eXecuted Latent Value Iteration Networks (XLVINs), which alleviate the above limitations. Our method performs all planning computations in a high-dimensional latent space, breaking the algorithmic bottleneck. It maintains alignment with value iteration by carefully leveraging neural graph-algorithmic reasoning and contrastive self-supervised learning. Across eight low-data settings -- including classical control, navigation and Atari -- XLVINs provide significant improvements to data efficiency against value iteration-based implicit planners, as well as relevant model-free baselines. Lastly, we empirically verify that XLVINs can closely align with value iteration.

arXiv commentTo appear at NeurIPS 2021 (Spotlight talk). 20 pages, 10 figures. arXiv admin note: text overlap with arXiv:2010.13146

References in corpus (11)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Recurrent World Models Facilitate Policy Evolution
• Learning Latent Dynamics for Planning from Pixels
• Differentiable MPC for End-to-end Planning and Control
• The Predictron: End-To-End Learning and Planning
• Learning Dynamic Belief Graphs to Generalize on Text-Based Games
• TreeQN and ATreeC: Differentiable Tree-Structured Models for Deep Reinforcement Learning
• Generalized Value Iteration Networks: Life Beyond Lattices
• On the role of planning in model-based deep reinforcement learning
• Graph neural induction of value iteration
• Reward Propagation Using Graph Convolutional Networks