# What Do Lorentz-Equivariant Jet Taggers Learn?

arXiv: https://arxiv.org/abs/2606.21790 · HTML: https://mlsanity.com/paper/2606.21790/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Circuits and mechanistic analysis, one of 13 directions in Mechanistic interpretability.
- Builds on 7 of its 13 references that are in the map.
- First posted on arXiv 19 Jun 2026.

## Placed in

- Mechanistic interpretability / [Circuits and mechanistic analysis](https://mlsanity.com/mechinterp/circuits-and-mechanistic-analysis/)

## Builds on (most cited first, 7 of 7)

- [Similarity of Neural Network Representations Revisited](https://mlsanity.com/paper/1905.00414/) (2019)
- [Particle Transformer for Jet Tagging](https://mlsanity.com/paper/2202.03772/) (2022)
- [Geometric Algebra Transformer](https://mlsanity.com/paper/2305.18415/) (2023)
- [Lorentz-Equivariant Geometric Algebra Transformers for High-Energy Physics](https://mlsanity.com/paper/2405.14806/) (2024)
- [Lorentz Local Canonicalization: How to Make Any Network Lorentz-Equivariant](https://mlsanity.com/paper/2505.20280/) (2025)
- [Interpretability Study on Deep Learning for Jet Physics at the Large Hadron Collider](https://mlsanity.com/paper/1911.01872/) (2019)
- [Interpreting Transformers for Jet Tagging](https://mlsanity.com/paper/2412.03673/) (2024)

## Abstract (arXiv)

We study what Lorentz-equivariant jet taggers learn internally, using equivariance tests, linear probes and grade ablations across five models including L-GATr, L-GATr-slim and LLoCa-T. Linear probes show that equivariant models suppress frame-dependent pseudorapidity to zero while encoding jet mass and N-subjettiness strongly. Grade ablations on L-GATr reveal that bivector channels are negligible for top-quark tagging while vector-like channels are dominant but seed variable, consistent with the network exploiting multiple representational pathways. These results characterize which physical features and algebraic grade structures carry discriminative information in equivariant taggers and may inform future development of such models.