Information based explanation methods for deep learning agents -- with applications on large open-source chess models
This work addresses the problem of interpretability in deep learning for chess, providing tools for researchers and practitioners using open-source models, though it is incremental in extending existing methods to new data.
The authors tackled the challenge of explaining domain knowledge in large chess-playing neural networks by re-implementing a concept detection method on open-source models and introducing a novel XAI method that guarantees exhaustive and exclusive highlighting of information used by the model, achieving results similar to those on proprietary models like AlphaZero.
With large chess-playing neural network models like AlphaZero contesting the state of the art within the world of computerised chess, two challenges present themselves: The question of how to explain the domain knowledge internalised by such models, and the problem that such models are not made openly available. This work presents the re-implementation of the concept detection methodology applied to AlphaZero in McGrath et al. (2022), by using large, open-source chess models with comparable performance. We obtain results similar to those achieved on AlphaZero, while relying solely on open-source resources. We also present a novel explainable AI (XAI) method, which is guaranteed to highlight exhaustively and exclusively the information used by the explained model. This method generates visual explanations tailored to domains characterised by discrete input spaces, as is the case for chess. Our presented method has the desirable property of controlling the information flow between any input vector and the given model, which in turn provides strict guarantees regarding what information is used by the trained model during inference. We demonstrate the viability of our method by applying it to standard 8x8 chess, using large open-source chess models.