The French National Institute of Nuclear and Particle Physics (French: Institut national de physique nucléaire et de physique des particules, IN2P3) is the coordinating body for nuclear and particle physics in France. It was established in 1971 as a division of the French National Centre for Scientific Research (CNRS). Its purpose is "to promote and unite research activities in the various fields of physics"..
We explore the connection of positivity of the imaginary part of forward elastic amplitudes for perturbative scattering with consistency of the entanglement generated by the S-matrix, for states with arbitrary internal quantum numbers such as flavor. We also analyze "disentanglers," certain highly entangled initial states for which the action of the S-matrix is to decrease subsystem entanglement. As a by-product, we develop a framework based on wave packets to regularize the spacetime divergences that appear in the plane-wave derivation of the entanglement expression.
We introduce a fast method to measure the conversion gain in complementary metal-oxide-semiconductor active pixel sensors, which accounts for nonlinearity and interpixel capacitance (IPC). The standard “mean-variance” method is biased because it assumes that pixel values depend linearly on the signal, and existing methods to correct for nonlinearity still introduce significant biases. While current IPC correction methods are prohibitively slow for a per-pixel application, our new method uses separate measurements of the IPC kernel to calculate the gain almost instantaneously. Using test data from a flight detector of the ESA Euclid mission, the IPC correction recovers the results of slower methods with 0.1% accuracy. The nonlinearity correction ensures that the estimated gain is independent of signal, correcting a bias of more than 2.5%.
We introduce a robust, interpretable machine learning (ML) framework that combines numerical regression for high-accuracy predictions with symbolic regression to uncover the underlying physics. This hybrid approach effciently derives analytical expressions by leveraging the smoothed predictions of optimized ML models, a significant acceleration over direct symbolic regression on raw experimental data. We apply this framework, as an example, to nuclear charge radii across the nuclear chart, notably including light nuclei that are often excluded from such studies. We employ Light Gradient Boosting Machine (LGBM) and Gaussian Process Regression (GPR) models to map correlations between charge radii and key physical features: mass A^1/3 and proton number Z^1/3 dependencies, total binding energy, and for the first time, the pairing gap. Our models are rigorously trained using four-fold cross-validation with automated hyperparameter optimization, ensuring robustness and generalizability, which is critical for the typically small and skewed datasets in nuclear physics. Finally, we distill the knowledge from the initial LGBM and GPR models into simplified, interpretable mathematical expressions via symbolic regression, white-boxing these ML models. The derived formulas provide physical insights comparable to traditional many-body models and demonstrate a powerful pathway for physics expression discovery guided by ML.
The phase-space approximation (PSA) approach, originally applied in [Phys. Rev. D 106, 123006 (2022)] to describe neutrino oscillations from a dense astrophysical source such as a core-collapse supernova in the two-flavor limit, is extended here to describe the more realistic case where neutrinos can oscillate between three different flavors. The approach is successfully validated against the exact solutions up to eight neutrinos. In all cases where the exact solution is feasible, the PSA provides excellent reproduction of the neutrino oscillation dynamics. By replacing the full problem with a set of simple mean-field equations, the PSA offers a versatile, predictive, and easily parallelizable approach for tackling three-flavor problems. This enables the simulation of large-scale neutrino oscillations, as illustrated here with simulations involving up to 300 neutrinos. Even if direct comparison to exact solutions is only feasible for small systems (up to eight neutrinos), where PSA shows excellent agreement, its applicability in the large-N regime appears promising, although a full quantitative validation remains beyond current computational capabilities. Additionally, the method also provides insight into the system's equilibration properties.
Production fraction ratios of B+, B-0, and B-s(0) mesons are measured in proton-proton collisions at root s = 13 TeV using a special dataset recorded in 2018 with high-rate triggers designed to collect an unbiased sample of 10(10) b hadrons with the CMS experiment at the LHC. These data allow the study of the opencharm decays of B mesons (B(s) -> pi D-(s) where the D meson decays into fully hadronic final states. By utilizing known branching fractions and precise theoretical calculations, production fraction ratios as functions of B meson transverse momentum (p(T)) and rapidity (y) are measured using open-charm decays in the kinematic range of 8 < p(T) < 60 GeV and vertical bar y vertical bar < 2.25. In addition, the same dataset is used to measure the relative production fraction ratios with the charmonium decay channels [B-(s) -> XJ/psi with X indicating a K+, K*(892)(0), or phi(1020) meson] where the J/psi meson decays into a pair of muons. The open-charm results are used to normalize the relative production fraction ratios obtained from the charmonium samples. Measurements of the ratios of branching fractions of B meson decays to charmonium and open-charm final states are also reported, which will improve the world-average values of these ratios. Finally, we test isospin invariance in B meson production in proton-proton collisions and observe that it holds within the experimental precision.
