HSF Seminar - HS3

Wednesday 25 Mar 2026, 16:30 → 18:00 Europe/Zurich

zoom

 

The HEP Software Foundation

The HEP Software Foundation facilitates cooperation and common efforts in High Energy Physics software and computing internationally. 
There are several focus activity areas: Data analysis, detector simulation, physics generators, PyHEP, Julia in HEP, reconstruction and software triggers, software developer tools and packaging, and training.
The HSF hosts workshops, provides strategic input and generally raises awareness on the importance of HEP software.

Sign up to HSF mailing lists here.

HSF Seminars

HSF Seminars were kick started in October 2024, and are nominally hosted on the last Wednesday of each month. Presentations are recorded (given consent) and put on YouTube.
See the full series on indico here!

Sign up to announcements for future HSF seminars as well as other HSF events by joining hsf-forum@googlegroups.com (sign up instructions).

Have a seminar topic idea? Send an email to hsf-seminar-conveners@googlegroups.com (current organisers are Nicole Skidmore, Michel Jouvin and Claire Antel)

 

16:30 → 16:35 Introduction

Speakers: Claire Antel (CERN), Dr Nicole Skidmore (University of Warwick)

16:35 → 17:20 From RooWorkspaces to Computational Graphs: HS3 and pyhs3

Statistical modeling is central to discovery in particle physics, yet the tools used to define, share, and evaluate these models are often fragmented or tightly coupled to legacy systems. The HEP Statistics Serialization Standard (HS³) addresses this with a framework-agnostic structure for serializing statistical models (distributions, functions, data, domains, parameter points, likelihoods, and analyses). Strongly motivated by the success of pyhf in enabling likelihood publishing and theory-side reinterpretation, HS³ extends this paradigm to the full scope of RooFit models and beyond. With growing adoption across ATLAS and CMS, a dedicated HEPData badge, integration into reinterpretation frameworks like CheckMATE, and an expanding interdisciplinary scope through the BMBF-funded DEMOS initiative, HS³ is rapidly becoming the community standard for FAIR statistical model exchange.

Additionally, this talk covers pyhs3, an ongoing pure-Python implementation that allows researchers to quickly get started working with HS³ workspaces without depending on ROOT or other legacy frameworks. Built on PyTensor, RustWorkX, SymPy, and Pydantic, pyhs3 constructs and evaluates computational graphs for likelihood ratio calculations with modularity, auto-differentiability, and integration with modern scientific computing ecosystems in mind. By providing a lightweight, well-documented entry point to the HS³ ecosystem, pyhs3 fills a key gap: enabling reinterpretation tools and theory workflows to consume published HS³ models directly, lowering the barrier for collaboration-independent reanalysis, likelihood combination, and EFT reparametrization. On-going work is done to validate its results against established calculations, examine performance relative to existing systems, and will outline future development plans.

Speaker: Dr Giordon Holtsberg Stark (University of California,Santa Cruz (US))