Search for lepton flavour violating τ->μμμ decays at Belle 2

• Robin LEBOUCHER ({CNRS}UMR7346)

The lepton flavour violating (LFV) decays are forbidden in the standard model (SM) of particle physics. Even by expanding the SM to explain neutrino oscillations, the τ LFV is extremely suppressed, and the discovery will be a clear mark of new physics beyond SM. The motivations for searching τ LFV have increased in the context of LHCb lepton flavour universality tests on b->sll and b->clν. In the Belle 2 experiment located at KEK in Japan, data taking started in 2019 to reach 400 1/fb before the 2023 long shutdown. Thanks to its clean environment and high ττ cross section, it provides an ideal environment to study the τ->μμμ decay, which is background free thanks to its purely leptonic final state. This analysis relies on an effective background suppression strategy presented here, with a preliminary look at data/simulation comparison in signal region sidebands.

Deep-learning data processing of spectral PC-CT longitudinal studies to design and optimize combined Immuno-anticancer Treatments in liver cancer mouse models

• Floriane Cannet

Hepatocellular carcinoma (HCC) is a late diagnosed cancer for which current treatments are not very effective, making it the third most deadly cancer in the world. In this context, we have developed at IBDM a unique mouse model (Alb-R26Met) that spontaneously generates HCC. The properties of these tumors allow to find all the heterogeneity and the resistance to treatment that we can have in humans. We used this mouse model to test the efficacy of a combination of treatments and to evaluate the response to treatment on these tumors. This response was evaluated by longitudinal follow-up made possible by the PIXSCAN-FLI, a photon counting spectral scanner (PC-CT) developed at the CPPM. The technology of detectors based on hybrid pixels makes it possible to carry out tomographies by counting photons and thus to obtain better qualities of images (less electronic noise, more contrast). Moreover, this technology also allows the development of spectral imaging at the K-edge whose benefits will be presented in this contribution.

Implementation of embeded artificial intelligence algorithms in the readout system of the ATLAS liquid argon calorimeter

• Nemer CHIEDDE ({CNRS}UMR7346)

The ATLAS experiment at CERN records proton-proton (p-p) collisions with a repetition frequency of 40MHz at the Large Hadron Collider (LHC). The readout electronics of liquid-argon (LAr) calorimeters are being prepared for high luminosity-LHC (HL-LHC) operation as part of the phase-II upgrade, anticipating a pileup of up to 200 simultaneous p-p interactions. The increase of the number of p-p interactions implies that calorimeter signals of up to 25 consecutive collisions overlap, making energy reconstruction more challenging. In order to achieve the goal of the HL-HLC, field-programmable gate arrays (FPGAs) are used to process digitized pulses sampled at 40MHz in real time and different machine learning approaches are being investigated to deal with signal pileup. The convolutional and recurrent neural networks outperform the optimal signal filter currently in use, both in terms of assigning the reconstructed energy to the correct proton bunch crossing and in terms of energy resolution. The enhancements are focused on energy obtained from overlapping pulses. Because the neural networks are implemented on an FPGA, the number of parameters, resource usage, latency and operation frequency must be carefully analyzed. A very good agreement is observed between neural network implementations in FPGA and software.