Interpretable and Explainable Logical Policies via Neurally Guided Symbolic Abstraction

About

The limited priors required by neural networks make them the dominating choice to encode and learn policies using reinforcement learning (RL). However, they are also black-boxes, making it hard to understand the agent's behaviour, especially when working on the image level. Therefore, neuro-symbolic RL aims at creating policies that are interpretable in the first place. Unfortunately, interpretability is not explainability. To achieve both, we introduce Neurally gUided Differentiable loGic policiEs (NUDGE). NUDGE exploits trained neural network-based agents to guide the search of candidate-weighted logic rules, then uses differentiable logic to train the logic agents. Our experimental evaluation demonstrates that NUDGE agents can induce interpretable and explainable policies while outperforming purely neural ones and showing good flexibility to environments of different initial states and problem sizes.

Quentin Delfosse, Hikaru Shindo, Devendra Dhami, Kristian Kersting · 2023

Related benchmarks

Task	Dataset	Result
Reinforcement Learning	Continuous Atari learning environment (CALE) DonkeyKong	Cumulative Reward21	13
Reinforcement Learning	Kangaroo Atari (classical)	Reward404	10
Reinforcement Learning	Seaquest Atari (classical)	Reward63	10
Reinforcement Learning	Kangaroo JAXAtari (test)	Score3058	6
Reinforcement Learning	Kingkong JAXAtari (test)	Final Score866	6
Reinforcement Learning	Seaquest JAXAtari (test)	Score64	6
Reinforcement Learning	OCAtari Asterix	Score62591150	3
Reinforcement Learning	OCAtari Freeway	Score—	2