Robust Independent Validation of Experiment and Theory: Rivet version 3

TL;DR

Rivet version 3 provides a robust, weight-aware framework for validating final-state predictions against data, unifying fiducial definitions, detector effects, and complex analyses across pp, heavy-ion, and ep collisions. It introduces a comprehensive system for multi-weight event handling, NLO counter-event treatment, and re-entrant run merging, enabling accurate statistical propagation and efficient analysis preservation. The update also extends Rivet to heavy-ion physics with centrality and flow formalisms, DIS/photoproduction projections, and practical detector-emulation wrappers, broadening its applicability to modern collider phenomenology and BSM searches. Together, these features enhance reproducibility, scalability, and interpretability of MC-data comparisons, while preserving a clear separation between analysis code and generator details. The result is a versatile, community-driven toolkit for fiducial, model-agnostic validation and reinterpretation of collider results.

Abstract

First released in 2010, the Rivet library forms an important repository for analysis code, facilitating comparisons between measurements of the final state in particle collisions and theoretical calculations of those final states. We give an overview of Rivet's current design and implementation, its uptake for analysis preservation and physics results, and summarise recent developments including propagation of MC systematic-uncertainty weights, heavy-ion and $ep$ physics, and systems for detector emulation. In addition, we provide a short user guide that supplements and updates the Rivet user manual.