Automated General Searches for New Physics

In contemporary high-energy physics, the number of potential scenarios for physics beyond the Standard Model far exceeds the capacity of dedicated searches. This disparity stresses the need for model-agnostic approaches that explore a broad spectrum of possibilities within the theoretical signal landscape. Developing effective methods for such searches requires optimizing the balance between generalizability and sensitivity while also addressing robustness, computational efficiency, and time demands.

Our focus is on designing strategies that refine these criteria while ensuring workflows are optimized for enhanced reproducibility and reinterpretability. Additionally, we aim to develop improved metrics for benchmarking model-agnostic search methods, particularly in terms of signal space coverage. The challenges posed by model-agnostic searches can be effectively addressed using cutting-edge machine-learning techniques, including advanced methods for reconstruction, pseudo-data generation, and signal classification. Beginning with theoretically well-motivated resonance searches over a smoothly falling background, these methods can potentially be extended to other scenarios, paving the way for broader and more efficient exploration of new physics.

Objectives:

• Develop methods for efficiently scanning a large signal phase space
• Validate methods for robustness and production of reliable results
• Characterise methods by universal, comparable metrics

Expected Results:

• Sensitivities of a robust and generalizable search to a large signal space
• Concrete application to a real physics dataset