# HEPTv2: End-to-End Efficient Point Transformer for Charged Particle Reconstruction

arXiv: https://arxiv.org/abs/2606.20437 · HTML: https://mlsanity.com/paper/2606.20437/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 43 references that are in the map.
- First posted on arXiv 18 Jun 2026.

## Builds on (most cited first, 24 of 24)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [End-to-End Object Detection with Transformers](https://mlsanity.com/paper/2005.12872/) (2020)
- [Masked-attention Mask Transformer for Universal Image Segmentation](https://mlsanity.com/paper/2112.01527/) (2021)
- [V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation](https://mlsanity.com/paper/1606.04797/) (2016)
- [Big Bird: Transformers for Longer Sequences](https://mlsanity.com/paper/2007.14062/) (2020)
- [Point Transformer V3: Simpler, Faster, Stronger](https://mlsanity.com/paper/2312.10035/) (2023)
- [Muon is Scalable for LLM Training](https://mlsanity.com/paper/2502.16982/) (2025)
- [Mask3D: Mask Transformer for 3D Semantic Instance Segmentation](https://mlsanity.com/paper/2210.03105/) (2022)
- [QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving](https://mlsanity.com/paper/2405.04532/) (2024)
- [FlatQuant: Flatness Matters for LLM Quantization](https://mlsanity.com/paper/2410.09426/) (2024)
- [FlatFormer: Flattened Window Attention for Efficient Point Cloud Transformer](https://mlsanity.com/paper/2301.08739/) (2023)
- [Learning representations of irregular particle-detector geometry with distance-weighted graph networks](https://mlsanity.com/paper/1902.07987/) (2019)
- [MLPF: Efficient machine-learned particle-flow reconstruction using graph neural networks](https://mlsanity.com/paper/2101.08578/) (2021)
- [ByteTransformer: A High-Performance Transformer Boosted for Variable-Length Inputs](https://mlsanity.com/paper/2210.03052/) (2022)
- [Performance of a Geometric Deep Learning Pipeline for HL-LHC Particle Tracking](https://mlsanity.com/paper/2103.06995/) (2021)
- [Charged particle tracking via edge-classifying interaction networks](https://mlsanity.com/paper/2103.16701/) (2021)
- [Graph Neural Networks for Charged Particle Tracking on FPGAs](https://mlsanity.com/paper/2112.02048/) (2021)
- [Scalable Deep Convolutional Neural Networks for Sparse, Locally Dense Liquid Argon Time Projection Chamber Data](https://mlsanity.com/paper/1903.05663/) (2019)
- [Locality-Sensitive Hashing-Based Efficient Point Transformer with Applications in High-Energy Physics](https://mlsanity.com/paper/2402.12535/) (2024)
- [High Pileup Particle Tracking with Object Condensation](https://mlsanity.com/paper/2312.03823/) (2023)
- [Hierarchical Graph Neural Networks for Particle Track Reconstruction](https://mlsanity.com/paper/2303.01640/) (2023)
- [GraphNeT: Graph neural networks for neutrino telescope event reconstruction](https://mlsanity.com/paper/2210.12194/) (2022)

## Abstract (arXiv)

Charged-particle tracking -- reconstructing trajectories from sparse detector measurements -- is a fundamental high-energy-physics inference problem and a canonical example of learning under extreme combinatorial ambiguity. At the High-Luminosity Large Hadron Collider (HL-LHC), tracking must remain accurate and efficient despite unprecedented collision densities. Graph neural networks perform strongly, but incur substantial costs from graph construction and processing, while transformer-based approaches rely on auxiliary stages that prevent end-to-end optimization. To address this, we present HEPTv2, an end-to-end point-transformer architecture that reconstructs tracks from detector hits in one trainable pipeline. HEPTv2 combines a locality-aware point encoder with a track decoder that predicts complete trajectories without graph-building, clustering, or filtering. The encoder uses locality-sensitive hashing in detector coordinate space to preserve tracking-relevant geometry while enabling efficient local attention. The decoder resolves ambiguities through sectorized decoding and direct hit-to-track prediction under joint encoder-decoder supervision, allowing the full pipeline to be optimized end-to-end. On TrackML, HEPTv2 achieves 98.6% double-majority tracking efficiency at a 0.8% fake rate, while requiring only ~15~ms inference time and 0.4~GB peak memory per event on a NVIDIA A100 GPU. Latency and memory scale approximately linearly for events with up to 5×105 hits. HEPTv2 establishes a new state of the art in the accuracy-latency trade-off, improving efficiency by 4.5% over the strongest prior transformer and by 1.1--2.2% over optimized graph-based pipelines, while reducing latency by factors of 7 and 38--52, respectively. These results show end-to-end transformers can deliver the accuracy and efficiency required for real-time particle reconstruction at the HL-LHC.