Fetching the paper…

Reading the bibliography…

2022

neos: End-to-End-Optimised Summary Statistics for High Energy Physics

Simpson, Nathan, Heinrich, Lukas

Original
PDF
Post on X

Understand

The advent of deep learning has yielded powerful tools to automatically compute gradients of computations.

This is because training a neural network equates to iteratively updating its parameters using gradient descent to find the minimum of a loss function.
Deep learning is then a subset of a broader paradigm; a workflow with free parameters that is end-to-end optimisable, provided one can keep track of the gradients all the way through.
This work introduces neos: an example implementation following this paradigm of a fully differentiable high-energy physics workflow, capable of optimising a learnable summary statistic with respect to the expected sensitivity of an analysis.

neos: End-to-End-Optimised Summary Statistics for High Energy Physics · Around

Built on

Hunter J D 2007 Computing in Science & Engineering
2007
Earlier work this paper cites.
Cowan G, Cranmer K, Gross E and Vitells O 2011 The European Physical Journal C
2011
Earlier work this paper cites.
ATLAS-Collaboration 2012 Physics Letters B
2012
Earlier work this paper cites.
Cranmer K, Lewis G, Moneta L, Shibata A and Verkerke W (ROOT Collaboration) 2012 HistFactory: A tool for creating statistical models for use with RooFit and RooStats Tech. rep. New York U. New York URL https://cds.cern.ch/record/1456844
2012
Earlier work this paper cites.

Similar

Louppe G, Kagan M and Cranmer K 2017 Learning to pivot with adversarial networks ( Preprint
2017
Cited alongside, same era.
de Castro P and Dorigo T 2019 Computer Physics Communications
2019
Cited alongside, same era.
Ghosh A, Nachman B and Whiteson D 2021 Physical Review D
2021
Cited alongside, same era.
Heinrich L, Feickert M, Stark G and Cranmer K 2021 Journal of Open Source Software
2021
Cited alongside, same era.
Heinrich L, Feickert M and Stark G pyhf: v0.6.3 https://github.com/scikit-hep/pyhf/releases/tag/v0.6.3 URL https://doi.org/10.5281/zenodo.1169739
Cited in the paper.
Duvenaud D, Johnson M and Kolter Z Deep implicit layers - neural odes, deep equilibirum models, and beyond URL http://implicit-layers-tutorial.org/
Cited in the paper.

Then

Blondel M, Berthet Q, Cuturi M, Frostig R, Hoyer S, Llinares-López F, Pedregosa F and Vert J P 2021 arXiv preprint arXiv:2105.15183
Original
2021
Later among the works it cites.
Simpson N and Heinrich L 2021 neos: version 0.2.0 URL https://github.com/gradhep/neos
2021
Later among the works it cites.
Simpson N 2022 relaxed: version 0.1.3 URL https://github.com/gradhep/relaxed
2022
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…