Economical Jet Taggers -- Equivariant, Slim, and Quantized
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910003538952192 |
|---|---|
| author | Petitjean, Antoine Plehn, Tilman Spinner, Jonas Köthe, Ullrich |
| author_facet | Petitjean, Antoine Plehn, Tilman Spinner, Jonas Köthe, Ullrich |
| contents | Modern machine learning is transforming jet tagging at the LHC, but the leading transformer architectures are large, not particularly fast, and training-intensive. We present a slim version of the L-GATr tagger, reduce the number of parameters of jet-tagging transformers, and quantize them. We compare different quantization methods for standard and Lorentz-equivariant transformers and estimate their gains in resource efficiency. We find an order-of-magnitude reduction in energy cost for an moderate performance decrease, down to 1000-parameter taggers. This might be a step towards trigger-level jet tagging with small and quantized versions of the leading equivariant transformer architectures. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_17011 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Economical Jet Taggers -- Equivariant, Slim, and Quantized Petitjean, Antoine Plehn, Tilman Spinner, Jonas Köthe, Ullrich High Energy Physics - Phenomenology High Energy Physics - Experiment Modern machine learning is transforming jet tagging at the LHC, but the leading transformer architectures are large, not particularly fast, and training-intensive. We present a slim version of the L-GATr tagger, reduce the number of parameters of jet-tagging transformers, and quantize them. We compare different quantization methods for standard and Lorentz-equivariant transformers and estimate their gains in resource efficiency. We find an order-of-magnitude reduction in energy cost for an moderate performance decrease, down to 1000-parameter taggers. This might be a step towards trigger-level jet tagging with small and quantized versions of the leading equivariant transformer architectures. |
| title | Economical Jet Taggers -- Equivariant, Slim, and Quantized |
| topic | High Energy Physics - Phenomenology High Energy Physics - Experiment |
| url | https://arxiv.org/abs/2512.17011 |