Explaining machine-learned particle-flow reconstruction
This work addresses the interpretability problem for physicists using machine-learned particle-flow algorithms, but it is incremental as it applies an existing technique to a specific domain.
The paper tackles the challenge of interpreting the complex decision-making of a graph neural network (GNN) used for particle-flow reconstruction in particle detectors, by adapting layerwise-relevance propagation to identify relevant nodes and features, resulting in insights into the model's predictions.
The particle-flow (PF) algorithm is used in general-purpose particle detectors to reconstruct a comprehensive particle-level view of the collision by combining information from different subdetectors. A graph neural network (GNN) model, known as the machine-learned particle-flow (MLPF) algorithm, has been developed to substitute the rule-based PF algorithm. However, understanding the model's decision making is not straightforward, especially given the complexity of the set-to-set prediction task, dynamic graph building, and message-passing steps. In this paper, we adapt the layerwise-relevance propagation technique for GNNs and apply it to the MLPF algorithm to gauge the relevant nodes and features for its predictions. Through this process, we gain insight into the model's decision-making.