Two measurements have been made, addressing gaps in knowledge for 235U(n,f) and 238U(n,f). The energy distribution for prompt fission neutrons is not well-understood below 1 MeV in 235U(n,f). To measure the 235U(n,f) prompt fission neutron distribution, a pulsed neutron beam at the WNR facility in Los Alamos National Laboratory was directed onto a 235U target with neutron detectors placed 1 m from the target. These neutron detectors were designed specifically for this experiment and employed a unique geometry of scintillating plastic material that was designed to reject backgrounds. Fission fragments were detected using an avalanche counter. Coincidences between fission fragment production and neutron detector events were analyzed, using a double time-of-flight technique to determine the energy of the prompt fission neutrons. A separate measurement was made, investigating the neutron-induced fission cross section for 238U(n,f). This measurement also used the pulsed neutron beam at the WNR facility. The neutron flux was normalized to the well-known hydrogen standard and the fission rate was observed for beam neutrons in the energy range of 130-300 MeV. Using an extrapolation technique, the energy dependence of the cross section was determined. These new data filled a sparsely populated energy region that was not well-studied and were measured relative to the hydrogen standard, unlike the majority of available data. These data can be used to constrain the fission cross section, which is considered a nuclear reaction standard.

This document, prepared to obtain the "Habilitation à Diriger des Recherches", is a compendium of the photon-related analysis activities I carried on within ATLAS in the past four years and a half. The activities I will describe can be broadly classified into three categories: optimization and/or in situ measurement of photon-related performance, measurements of the cross sections of Standard Model processes producing prompt photons, and searches (leading to discovery!) of a Standard Model Higgs boson decaying to final states containing photons.

This work is aimed at the understanding of the dynamical behavior of hydrogen in diamond. The investigation was carried out using Transverse Field muon Spin Rotation (TF-SR) and the Longitudinal Field muon Spin Relaxation (LF-SR) techniques. The chemical analogy between hydrogen (p+e-) and muonium (u+e-) enabled the study of the indirect dynamical behavior of hydrogen in diamond. The TF-SR and LF-SR measurements were carried out in an isotopically pure 13C diamond in the temperature ranges 11 mK - 320 K and 10 K - 400 K, respectively. In the TF-SR results, the Prompt Absolute Fraction (PAF) of both diamagnetic and the paramagnetic states are temperature independent. The spin relaxation rate for the state is non-zero and temperature independent, while that of the Mux state is non-zero and temperature dependent. The behavior of indicates that the u+d is immobile, while that of MuT indicates the mobility in diamond. The LF-SR results provide hop rate and associated nuclear hyperfine interaction parameters of the Mut state in diamond. / Physics / M.Sc. (Physics)

L'hadronthérapie est une technique innovante de radiothérapie par ions carbone ou protons visant à améliorer les traitements actuels. La précision balistique accrue renforce la nécessité d'un contrôle du dépôt de dose, si possible en temps réel. Une manière de réaliser ce contrôle est de détecter avec une caméra Compton le gamma prompt émis lors des fragmentations nucléaires pendant l'irradiation du patient. Dans un premier temps, la géométrie de deux types de caméra Compton (double diffusion puis simple diffusion) a été optimisée par simulation Monte Carlo. Cette optimisation a été réalisée en étudiant la réponse des caméras à une source ponctuelle de photons avec un spectre en énergie réaliste. La réponse de la caméra optimisée à l'irradiation d'un fantôme d'eau par un faisceau d'ions carbone ou de protons a ensuite été simulée. Ces simulations ont tout d'abord été confrontées à des mesures effectuées avec un prototype de taille réduite. Ces mesures ont ensuite été utilisées pour évaluer les taux de comptage dans les détecteurs attendus en conditions cliniques. Dans la configuration actuelle de la caméra, ces taux sont élevés et les phénomènes d'empilement risquent d'être problématiques. Enfin, il est démontré que le dispositif étudié est sensible à un déplacement du pic de Bragg de plus ou moins 5 mm malgré les problèmes de coïncidences fortuites et malgré le bruit introduit par l'algorithme de reconstruction utilisé. / Ion beam therapy is an innovative radiotherapy technique using mainly carbon ion and proton irradiations. Its aim is to improve the current treatment modalities. Because of the sharpness of the dose distributions, a control of the dose if possible in real time is highly desirable. A possibility is to detect the prompt gamma rays emitted subsequently to the nuclear fragmentations occurring during the treatment of the patient. In a first time two different Compton cameras (double and single scattering) have been optimised by means of Monte Carlo simulations. The response of the camera to a photon point source with a realistic energy spectrum was studied. Then, the response of the camera to the irradiation of a water phantom by a proton beam was simulated. It was first compared with measurement performed with small-size detectors. Then, using the previous measurements, we evaluated the counting rates expected in clinical conditions. In the current set-up of the camera, these counting rates are pretty high. Pile up and random coincidences will be problematic. Finally we demonstrate that the detection system is capable to detect a longitudinal shift in the Bragg peak of +or- 5 mm, even with the current reconstruction algorithm.

The IceCube Neutrino Observatory is a cubic kilometre scale detector for high-energy neutrinos above hundreds of GeV produced in Earth’s atmosphere as well as outside our solar system whenever particles are accelerated to ultra-relativistic energies. The prompt atmospheric contribution is a result of the creation of heavy mesons with charm components in the atmosphere. Past studies from IceCube using a maximum likelihood estimation over the whole neutrino energy spectrum always reported a best-fit zero prompt contribution so far [1–5], contrary to theory [6, 7]. In this analysis we tried to measure this prompt atmospheric flux in muon neutrino event data from different IceCube releases. In contrast to past studies we performed a binned least-squares fit of the conventional atmospheric flux from data at low energies and subtracted this fit and an astrophysical flux reported by IceCube to measure a prompt contribution. Due to a lack of statistics and accessible information from data releases, our results are also compatible with a zero prompt contribution.

La fission nucléaire est un phénomène complexe dont tous les mécanismes ne sont pas entièrement compris. Dans le cadre d'une coopération internationale, le CEA/DAM/DIF étudie les spectres en énergie des neutrons prompts émis lors de la fission induite par des neutrons rapides, et plus particulièrement la zone à basse énergie de ces spectres (<1 MeV). Ce travail de thèse a consisté à optimiser un dispositif expérimental de mesure de neutrons prompts de fission. Dans un premier temps, de nouveaux détecteurs de fission ont été développés. Nous en rapportons ici la conception et étudions leurs performances en termes de discrimination alpha-fission, de résolution en temps et de distorsion sur le spectre mesuré. Le second axe de développement abordé au cours de cette thèse est celui de la détection des neutrons. Plusieurs types de détecteurs ont été comparés (discrimination neutron-gamma, efficacité de détection), en vue d'optimiser la détection des neutrons de basse énergie (<1 MeV). Ce mémoire présente les résultats de ces études. Enfin, le dispositif expérimental ainsi optimisé est utilisé pour mesurer le spectre en énergie des neutrons prompts émis lors de la fission induite par neutrons de l' ²³⁸UU. Après avoir présenté la méthode utilisée pour l'analyse des données, les résultats obtenus sont interprétés en termes de modèles et d'évaluations. / The nuclear fission is a complex phenomenon whose mechanisms are not fully understood. Within the framework of an international cooperation, the CEA/DAM/DIF is taking part in the study of prompt fission neutron energy spectra from fast neutron induced fission, focusing on the low energy domain of these spectra (<1 MeV). This PhD was dedicated to the optimization of the experimental setup. New fission detectors were developed. We report on their conception and their performances in terms of alpha-fission discrimination, timing resolution and distortion on the measured spectrum. In a second step, several neutron detectors were studied (neutron-gamma discrimination, detection efficiency), so as to optimize the detection of low energy neutrons (<1 MeV). In the present document, we report on the results of this comparative study. Finally, the optimized experimental setup was used to measure prompt fission neutron energy spectra for the fast-neutron induced fission of ²³⁸U. After detailing the data analysis method, the results are interpreted in terms of models and evaluations.

<p>The present dissertation investigated the impact of scales / scoring methods and prompt linguistic features on the meausrement quality of L2 English elicited imitation (EI). Scales / scoring methods are an important feature for the validity and reliabilty of L2 EI test, but less is known (Yan et al., 2016). Prompt linguistic features are also known to influence EI test quaity, particularly item difficulty, but item discrimination or corpus-based, fine-grained meausres have rarely been incorporated into examining the contribution of prompt linguistic features. The current study addressed the research needs, using item response theory (IRT) and random forest modeling.</p><p>Data consisted of 9,348 oral responses to forty-eight items, including EI prompts, item scores, and rater comments, which were collected from 779 examinees of an L2 English EI test at Purdue Universtiy. First, the study explored the current and alternative EI scales / scoring methods that measure grammatical / semantic accuracy, focusing on optimal IRT-based measurement qualities (RQ1 through RQ4 in Phase Ⅰ). Next, the project identified important prompt linguistic features that predict EI item difficulty and discrimination across different scales / scoring methods and proficiency, using multi-level modeling and random forest regression (RQ5 and RQ6 in Phase Ⅱ).</p><p>The main findings were (although not limited to): 1) collapsing exact repetition and paraphrase categories led to more optimal measurement (i.e., adequacy of item parameter values, category functioning, and model / item / person fit) (RQ1); there were fewer misfitting persons with lower proficiency and higher frequency of unexpected responses in the extreme categories (RQ2); the inconsistency of qualitatively distinguishing semantic errors and the wide range of grammatical accuracy in the minor error category contributed to misfit (RQ3); a quantity-based, 4-category ordinal scale outperformed quality-based or binary scales (RQ4); sentence length significantly explained item difficulty only, with small variance explained (RQ5); Corpus-based lexical measures and phrase-level syntactic complexity were important to predicting item difficulty, particularly for the higher ability level. The findings made implications for EI scale / item development in human and automatic scoring settings and L2 English proficiency development.</p>

Behavioral skills training (BST) is often used to train caregivers to implement various behavior management procedures; however, additional strategies are sometimes required to promote the generalization of skills from a contrived setting to the natural environment. Generalizing skills to the natural environment requires that the caregiver's behavior transfer from control of stimuli in the contrived setting to stimuli in the natural environment, and the skill continues to be performed with high levels of accuracy. The purpose of this study was to assess the extent to which caregivers generalized the use of social reinforcement, in the form of descriptive praise, from the contrived setting to the natural environment. When caregivers failed to respond to opportunities, a progressive prompt delay was used to bring caregivers' responding under the control of relevant client behavior; feedback was used to improve the accuracy with which caregivers implemented reinforcement. Five caregivers in a large residential facility participated in the study; single-opportunity probes were used to assess caregiver's identification of opportunities and accuracy in implementing reinforcement for two defined client behaviors, compliance and appropriate attention-getting behavior. Results of the study suggest that skills failed to generalize from the contrived setting to the natural environment. However, prompting was effective in training caregivers to identify opportunities to provide reinforcement, and feedback improved implementation of reinforcement.

<p>Empowering machines to understand our physical world should go beyond models with only natural language and models with only vision. Vision and language is a growing field of study that attempts to bridge the gap between natural language processing and computer vision communities by enabling models to learn visually grounded language. However, as an increasing number of pre-trained visual linguistic models focus on the alignment between visual regions and natural language, it is difficult to claim that these models capture certain properties of objects in their latent space, such as size. Inspired by recent trends in prompt learning, this study will design a prompt learning framework for two visual linguistic models, ViLBERT and ViLT, and use different manually crafted prompt templates to evaluate the consistency of performance of these models in comparing the size of objects. The results of this study showed that ViLT is more consistent in prediction accuracy for the given task with six pairs of objects under four prompt designs. However, the overall prediction accuracy is lower than the expectation on this object size comparison task; even the better model in this study, ViLT, has only 16 out of 24 cases better than the proposed random chance baseline. As this study is a preliminary study to explore the potential of pre-trained visual linguistic models on object size comparison, there are many directions for future work, such as investigating more models, choosing more object pairs, and trying different methods for feature engineering and prompt engineering.</p>

The knowledge of tests conducted on a problematic vehicle is essential for enhancing the efficiency of mechanics. Therefore, identifying the tests performed in each repair case is of utmost importance. This thesis explores techniques for extracting data from unstructured repair reports to identify component tests. The main emphasis is on developing a supervised multi-class classifier to categorize data and extract sentences that describe repair diagnoses and actions. It has been shown that incorporating a category-aware contrastive learning objective can improve the repair report classifier’s performance. The proposed approach involves training a sentence representation model based on a pre-trained model using a category-aware contrastive learning objective. Subsequently, the sentence representation model is further trained on the classification task using a loss function that combines the cross-entropy and supervised contrastive learning losses. By applying this method, the macro F1-score on the test set is increased from 90.45 to 90.73. The attempt to enhance the performance of the repair report classifier using a noisy data classifier proves unsuccessful. The noisy data classifier is trained using a prompt-based fine-tuning method, incorporating open-ended questions and two examples in the prompt. This approach achieves an F1-score of 91.09 and the resulting repair report classification datasets are found easier to classify. However, they do not contribute to an improvement in the repair report classifier’s performance. Ultimately, the repair report classifier is utilized to aid in creating the input necessary for identifying component tests. An information retrieval method is used to conduct the test identification. The incorporation of this classifier and the existing labels when creating queries leads to an improvement in the mean average precision at the top 3, 5, and 10 positions by 0.62, 0.81, and 0.35, respectively, although with a slight decrease of 0.14 at the top 1 position.