Leading-colour-based unweighted event generation for multi-parton tree-level processes
Rikkert Frederix, Timea Vitos

TL;DR
This paper presents a two-step method for efficient unweighted event generation in multi-parton QCD processes, using leading-colour accuracy followed by reweighting to full-colour accuracy, significantly improving efficiency.
Contribution
The paper introduces a novel two-step approach that combines leading-colour event generation with reweighting to full-colour accuracy, enhancing efficiency in multi-parton process simulations.
Findings
High primary unweighting efficiencies (percent to per-mille level) for 2→4 to 2→7 processes.
Secondary unweighting efficiencies exceed 50% due to close approximation between LC and FC matrix elements.
The method offers an efficient alternative to direct full-colour event generation.
Abstract
In this work, we revisit unweighted event generation for multi-parton tree-level processes in massless QCD. We introduce a two-step approach, in which initially unweighted events are generated at leading-colour (LC) accuracy, followed by a reweighting of these events to full-colour (FC) accuracy and applying an additional unweighting cycle. This method leverages the simple structure of LC integrands, enabling optimized phase-space parameterisations and resulting in high primary unweighting efficiencies, ranging from the percent level for processes to the per-mille level for processes. Given that the LC-accurate matrix elements closely approximate the FC-accurate ones, the secondary unweighting efficiencies exceed 50%. Our results suggest that this two-step approach offers an efficient alternative to direct event generation at FC accuracy.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Data Storage Technologies · Simulation Techniques and Applications · Scientific Computing and Data Management
