Related papers: LINX: A Fast, Differentiable, and Extensible Big B…

We present the results from combining machine learning with the profile likelihood fit procedure, using data from the Large Underground Xenon (LUX) dark matter experiment. This approach demonstrates reduction in computation time by a factor…

Standard Big Bang Nucleosynthesis (BBN) predicts the abundances of the light elements in the early universe. Even if the overall agreement with the experimental data is good, still some discrepancies exist on the relic abundances of…

In this paper we investigate XSB-Prolog as a static analysis engine for data represented by medium-sized graphs. We use XSB-Prolog to automatically identify function dependencies in the Linux Kernel---queries that are difficult to implement…

Combined with other CMB experiments, the WMAP survey provides an accurate estimate of the baryon density of the Universe. In the framework of the standard Big Bang Nucleosynthesis (BBN), such a baryon density leads to predictions for the…

xBIT is a tool for performing parameter scans in beyond the Standard Model theories. It's written in Python and fully open source. The main purpose of xBIT is to provide an easy to use tool to help phenomenologists with their daily task:…

Big Bang nucleosynthesis (BBN) is the earliest sensitive probe of the values of many fundamental particle physics parameters. We have found the leading linear dependences of primordial abundances on all relevant parameters of the standard…

We propose local binary convolution (LBC), an efficient alternative to convolutional layers in standard convolutional neural networks (CNN). The design principles of LBC are motivated by local binary patterns (LBP). The LBC layer comprises…

Primordial or big bang nucleosynthesis (BBN) is now a parameter free theory whose predictions are in good overall agreement with observations. However, the 7Li calculated abundance is significantly higher than the one deduced from…

Large language models (LLMs) are increasingly touted as powerful tools for automating scientific information extraction. However, existing methods and tools often struggle with the realities of scientific literature: long-context documents,…

The Framework for Unified and Robust data Analysis with JAX (Furax) is an open-source Python framework for modeling data acquisition systems and solving inverse problems in astrophysics and cosmology. Built on JAX, Furax provides composable…

From the observations of the anisotropies of the Cosmic Microwave Background (CMB) radiation, the WMAP satellite has provided a determination of the baryonic density of the Universe, \Omega_b.h^2, with an unprecedented precision. This…

Intermediate-layer predictions in large language models (LLMs) are informative but hard to decode accurately, especially at early layers. Existing lens-style methods typically rely on direct linear readout, which is simple but often drifts…

We introduce Atomistic learned potentials in JAX (apax), a flexible and efficient open source software package for training and inference of machine-learned interatomic potentials. Built on the JAX framework, apax supports GPU acceleration…

We introduce fastabx, a high-performance Python library for building ABX discrimination tasks. ABX is a measure of the separation between generic categories of interest. It has been used extensively to evaluate phonetic discriminability in…

Big Bang nucleosynthesis in the presence of decaying relic neutral particles is examined in detail. All non-thermal processes important for the determination of light-element abundance yields of 2H, 3H, 3He, 4He, 6Li, and 7Li are coupled to…

Primordial or Big Bang nucleosynthesis (BBN) is one of the three strong evidences for the Big- Bang model together with the expansion of the Universe and the Cosmic Microwave Background radiation. In this study, we improve the standard BBN…

We advocate for a new paradigm of cosmological likelihood-based inference, leveraging recent developments in machine learning and its underlying technology, to accelerate Bayesian inference in high-dimensional settings. Specifically, we…

The key Standard-Physics inputs of the Big Bang Nucleosynthesis (BBN) are the light nuclei reaction rates. Both the network and the nuclear rates have been recently reanalyzed and updated, and cosmological and New-Physics constraints…

This paper reviews the state of art technology of laser induced breakdown spectroscopy (LIBS). Research on LIBS is gaining momentum in the field of instrumentation and data analysis technique due to its wide application in various field…

This work proposes a spectral convolutional neural network (CNN) operating on laser induced breakdown spectroscopy (LIBS) signals to learn to (1) disentangle spectral signals from the sources of sensor uncertainty (i.e., pre-process) and…