# QCD-Aware Recursive Neural Networks for Jet Physics

## TL;DR

This paper introduces a novel recursive neural network approach for jet physics that models jets as language-like structures, improving accuracy and data efficiency over previous image-based methods.

## Contribution

It presents a new QCD-aware recursive neural network architecture that directly processes particle four-momenta and jet clustering trees, extending to event-level classification.

## Key findings

- Recursive networks outperform image-based models in accuracy.
- Method is more data-efficient.
- Extends to event-level classification.

## Abstract

Recent progress in applying machine learning for jet physics has been built upon an analogy between calorimeters and images. In this work, we present a novel class of recursive neural networks built instead upon an analogy between QCD and natural languages. In the analogy, four-momenta are like words and the clustering history of sequential recombination jet algorithms is like the parsing of a sentence. Our approach works directly with the four-momenta of a variable-length set of particles, and the jet-based tree structure varies on an event-by-event basis. Our experiments highlight the flexibility of our method for building task-specific jet embeddings and show that recursive architectures are significantly more accurate and data efficient than previous image-based networks. We extend the analogy from individual jets (sentences) to full events (paragraphs), and show for the first time an event-level classifier operating on all the stable particles produced in an LHC event.