- Feb. 17, 2022, 2:30 pm US/Central
- Elias Bernreuther, FNAL
- video
Searches for missing energy at accelerators are a central pillar of the experimental program devoted to solving the dark matter puzzle. However, extended dark sectors may manifest themselves in the form of exotic signatures that are missed by standard searches. In my talk, I will discuss a particularly striking class of such signatures: dark showers originating from strongly interacting dark sectors. I will argue that such sectors can be at the same time cosmologically viable and in reach of current and future LHC searches. I will then show that the sensitivity of these searches can be improved substantially with machine learning, focusing on tagging the substructure of semi-visible jets with dynamic graph convolutional neural networks. Furthermore, strongly interacting dark sectors motivate searches for light long-lived particles at the LHC and at the intensity frontier. I will show in particular that the Belle II experiment has excellent sensitivity for dark showers consisting of long-lived dark mesons with mass on the sub-GeV scale.