Researchers have developed a novel application of reinforcement learning (RL) for real-time event filtering at the Large Hadron Collider (LHC). By adapting Group-Filtered Policy Optimization (GFPO) to streaming data, the RL agent can dynamically adjust trigger thresholds to maximize signal efficiency while maintaining a target background rate. This approach demonstrated significant improvements, increasing in-tolerance time by 56% for a transverse energy trigger and 28% for an anomaly detection trigger when applied to real CMS Run 283408 collision data, marking the first use of RL for trigger control in such an environment. AI
IMPACT Demonstrates a new method for optimizing complex real-time systems using AI, potentially applicable to other high-throughput scientific facilities.
RANK_REASON Academic paper detailing a novel application of reinforcement learning to a scientific instrument. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Hugging Face Daily Papers →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
