arXiv subjects

Tyler Wheeler

Publications and source records attributed to Tyler Wheeler.

4 recordsLinked to original sources

arXiv2026

How Architecture and Training Affect TPC Representations Across Experiments

Deep-learning efforts have increasingly shifted toward foundation model approaches. In experimental physics, this allows models and learned representations to be reused beyond the experiments in which they were developed. This work evaluates the reusability of representations across experiments and detector systems using probes on frozen encoders. These probes reveal task-relevant structure before downstream adaptation, complementing fine-tuning. Together with random-weight controls, they distinguish contributions from architecture and encoder training that downstream performance alone cannot resolve. Time projection chamber (TPC) data provide a useful testbed because events from TPC systems can be represented as variable-length sparse tensors, while detector geometries, event topologies, and scientific tasks can differ substantially. We investigate whether fixed-dimensional TPC event representations can be reused across classification tasks, experiments, and detector systems. Sparse ResNet and PointNet-style encoders produce 512-dimensional embeddings for four datasets from the GADGET II TPC and AT-TPC. Randomly initialized encoders isolate the contribution from architecture before supervised training. We then train each encoder on a classification task, freeze its parameters, and train a linear or nonlinear probe for each downstream task. We find that this architecture-induced structure remains useful across experiments and detector systems. The randomly initialized PointNet-style representation is highly informative on several tasks. The two architectures organize their embedding spaces differently, but neither exhibits a large, systematic loss of utility cross-detector. These results show that architecture is a major source of task-relevant structure in TPC embeddings and should be treated explicitly when assessing representation learning and developing reusable detector models.

cs.LG↗

arXiv2025

Sparse Methods for Vector Embeddings of TPC Data

Time Projection Chambers (TPCs) are versatile detectors that reconstruct charged-particle tracks in an ionizing medium, enabling sensitive measurements across a wide range of nuclear physics experiments. We explore sparse convolutional networks for representation learning on TPC data, finding that a sparse ResNet architecture, even with randomly set weights, provides useful structured vector embeddings of events. Pre-training this architecture on a simple physics-motivated binary classification task further improves the embedding quality. Using data from the GAseous Detector with GErmanium Tagging (GADGET) II TPC, a detector optimized for measuring low-energy $\beta$-delayed particle decays, we represent raw pad-level signals as sparse tensors, train Minkowski Engine ResNet models, and probe the resulting event-level embeddings which reveal rich event structure. As a cross-detector test, we embed data from the Active-Target TPC (AT-TPC) -- a detector designed for nuclear reaction studies in inverse kinematics -- using the same encoder. We find that even an untrained sparse ResNet model provides useful embeddings of AT-TPC data, and we observe improvements when the model is trained on GADGET data. Together, these results highlight the potential of sparse convolutional techniques as a general tool for representation learning in diverse TPC experiments.

cs.LG↗

arXiv2025

$\beta$-delayed proton pandemonium: A first detailed $^{31}$Cl($\beta p \gamma$)$^{30}$P decay scheme

Positron decays of proton-rich nuclides exhibit large $Q$ values, producing complex cascades which frequently involve various radiations, including protons and $\gamma$ rays. Often, only one of the two is measured in a single experiment, limiting the accuracy and completeness of the decay scheme. An example is $^{31}$Cl, for which protons and $\gamma$ rays have been measured separately in detail but never with substantial sensitivity to proton-$\gamma$ coincidences. We provide detailed measurements of $^{31}$Cl $\beta$-delayed proton decay including $\beta$-$p$-$\gamma$ sequences, extract spectroscopic information on $^{31}$S excited states as well as their $\beta^+$ feedings, and compare to shell-model calculations. A fast fragmented beam of $^{31}$Cl provided by the National Superconducting Cyclotron Laboratory (NSCL) was deposited in the Gaseous Detector with Germanium Tagging (GADGET) system. GADGET's gas-filled Proton Detector was used to detect $\beta$-delayed protons, and the Segmented Germanium Array (SeGA) was used to detect $\beta$-delayed $\gamma$ rays. As many as 20 previously unobserved $\beta$-delayed proton transitions are reported, most of which populate excited states of $^{30}$P. The first detailed $^{31}$Cl($\beta p \gamma$)$^{30}$P decay scheme is presented, including updated $\beta$-delayed proton energies and intensities, as well as several new $^{31}$S levels. Improved agreement is found with theoretical calculations of the Gamow-Teller strengths $B(\text{GT})$ for $^{31}$S excitation energies $7.5 < E_x < 9.5$ MeV. The present work demonstrates that the ability to detect $\beta$-delayed protons and $\gamma$ rays in coincidence is essential for accurate positron decay schemes to compare with nuclear structure theory. This phenomenon for $\beta$-delayed protons resembles the pandemonium effect originally introduced for $\beta$-delayed $\gamma$ rays.

nucl-ex↗

arXiv2025

Object Detection with Deep Learning for Rare Event Search in the GADGET II TPC

In the pursuit of identifying rare two-particle events within the GADGET II Time Projection Chamber (TPC), this paper presents a comprehensive approach for leveraging Convolutional Neural Networks (CNNs) and various data processing methods. To address the inherent complexities of 3D TPC track reconstructions, the data is expressed in 2D projections and 1D quantities. This approach capitalizes on the diverse data modalities of the TPC, allowing for the efficient representation of the distinct features of the 3D events, with no loss in topology uniqueness. Additionally, it leverages the computational efficiency of 2D CNNs and benefits from the extensive availability of pre-trained models. Given the scarcity of real training data for the rare events of interest, simulated events are used to train the models to detect real events. To account for potential distribution shifts when predominantly depending on simulations, significant perturbations are embedded within the simulations. This produces a broad parameter space that works to account for potential physics parameter and detector response variations and uncertainties. These parameter-varied simulations are used to train sensitive 2D CNN object detectors. When combined with 1D histogram peak detection algorithms, this multi-modal detection framework is highly adept at identifying rare, two-particle events in data taken during experiment 21072 at the Facility for Rare Isotope Beams (FRIB), demonstrating a 100% recall for events of interest. We present the methods and outcomes of our investigation and discuss the potential future applications of these techniques.

physics.ins-det↗