Economical Jet Taggers -- Equivariant, Slim, and Quantized

Abstract

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.

v1: 21 pages, 7 tables, 6 figures; v2: 23 pages, 6 figures, 8 tables, add quantized ParT, int8 instead of float8 quantization, refine text