SciPost Submission Page
Towards Universal Unfolding of Detector Effects in High-Energy Physics using Denoising Diffusion Probabilistic Models
by Camila Pazos, Shuchin Aeron, Pierre-Hugues Beauchemin, Vincent Croft, Zhengyan Huan, Martin Klassen, Taritree Wongjirad
Submission summary
| Authors (as registered SciPost users): | Camila Pazos |
| Submission information | |
|---|---|
| Preprint Link: | scipost_202410_00060v2 (pdf) |
| Code repository: | https://gitlab.cern.ch/cpazosbo/cddpm-unfolder |
| Data repository: | https://zenodo.org/records/13993067 |
| Date accepted: | June 12, 2025 |
| Date submitted: | June 9, 2025, 12:43 a.m. |
| Submitted by: | Pazos, Camila |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approaches: | Experimental, Computational |
Abstract
Correcting for detector effects in experimental data, particularly through unfolding, is critical for enabling precision measurements in high-energy physics. However, traditional unfolding methods face challenges in scalability, flexibility, and dependence on simulations. We introduce a novel approach to multidimensional object-wise unfolding using conditional Denoising Diffusion Probabilistic Models (cDDPM). Our method utilizes the cDDPM for a non-iterative, flexible posterior sampling approach, incorporating distribution moments as conditioning information, which exhibits a strong inductive bias that allows it to generalize to unseen physics processes without explicitly assuming the underlying distribution. Our results highlight the potential of this method as a step towards a "universal" unfolding tool that reduces dependence on truth-level assumptions, while enabling the unfolding of a wide range of measured distributions with improved adaptability and accuracy.
Author indications on fulfilling journal expectations
- Provide a novel and synergetic link between different research areas.
- Open a new pathway in an existing or a new research direction, with clear potential for multi-pronged follow-up work
- Detail a groundbreaking theoretical/experimental/computational discovery
- Present a breakthrough on a previously-identified and long-standing research stumbling block
List of changes
Content and Structure Changes - Revised "Our Contribution" section to emphasize moment-conditioning as the main novelty rather than just using a different generative network - Enhanced Section 2.1 and 2.2 with clarifying sentences linking moment-conditioning to specific generative model requirements - Added paragraph at end of Section 2.2 comparing cDDPM approach to other diffusion-based unfolding methods - Clarified evaluation procedure in Section 2.3 for how moments are calculated on real experimental data - Added lines to Section 2.3 clarifying that Figure 4 demonstrates the importance of moment conditioning vs. diverse training data - Added paragraph in Conclusion addressing lack of direct quantitative comparisons to other ML methods and emphasizing moment-conditioning novelty - Added reference to the OmniFold training data augmentation study (2105.09923) in Related Works section - Moved Figure 1 from Introduction to Section 3 to improve flow
Figure Improvements - Increased font sizes in all figures for better readability - Converted three-panel figures to two-panel (Figures 3, 5, 6) for clarity
Current status:
Editorial decision:
For Journal SciPost Physics Core: Publish
(status: Editorial decision fixed and (if required) accepted by authors)