Projects | Oz Amram

Anomaly Detection

Model-agnostic searches that let the data itself flag new phenomena, broadening our sensitivity to the unexpected.

The first application of anomaly detection in CMS

A new method for anomaly detection at the LHC

Unearthing a new class of anomaly detection signatures

Using diffusion models and normalizing flows to accelerate detector simulation and to perform high-dimensional, data-driven inference.

Diffusion models for LHC calorimeter simulations

Data-driven high-dimensional inference with generative models

Benchmarking generative models as fast simulation for the CMS calorimeter upgrade

Training large models on real LHC collision data, as a basis for many downstream physics tasks.

Pre-training on real LHC data, and how physics data scales

Earlier work on detector calibration and precision electroweak measurements.

Measuring the Drell-Yan forward-backward asymmetry with CMS

Calibration and Reconstruction