My latest paper is on the arXiv today, to be submitted to JHEP (Journal of High Energy Physics): https://arxiv.org/abs/2203.01009
Here is a simple explanation…
The Standard Model of particle physics encapsulates our best understanding of the building blocks of the universe to date, but we know it’s incomplete: there must ne new physics to account for dark matter. So far, no extra fundamental particles have been discovered at the LHC’s two main detectors, which can be thought of as 3D cameras surrounding the proton collision points. But what if we were looking in the wrong place? So far, we have assumed that any new particles would instantly decay at the collision point, before they reach our detectors… but particles which travelled a few meters and decayed **inside** the detector itself would have been completely missed. Just like in a car, it’s crucial to check the blind spots, in case something important is hiding there.
So this paper is part of a series of searches for so called “long-lived” exotic particles. We used AI to try to distinguish candidate collision events which might contain these particles decaying “inside the camera”, from some of the many unusual backgrounds we face: the ATLAS detectors were not designed for this sort of search! Sadly we did not find any evidence of new particles, but we set some stringent limits on how often any such particles could be produced at the scale of the LHC. The search goes on!
Here are more explanations about the picture: neutral long-lived particles decaying inside the ATLAS calorimeters would lead to narrow, trackless jets (clusters of energy deposits) with little or no energy in electromagnetic part of the calorimeter. You see two such jets in the picture, as bunches of yellow bars, with very few yellow tracks in their vicinity. This collision event is from actual LHC data (it’s not a simulation, just a visualisation), and it looks just like our signal (the long-lived particles) would. However, regular collisions can also cause such events, because of inefficiencies in detector reconstruction, and we precisely measured how many of these “background” collisions we would expect. We didn’t observe any excess of such events above what we expected from the background, so it’s likely this particular event was **not** due to new physics!