Search

We present a coil system designed to generate a highly uniform magnetic field for the n2EDM experiment at the Paul Scherrer Institute. It consists of a main $B_0$ coil and a set of auxiliary coils mounted on a cubic structure with a side length of 273 cm, inside a large magnetically shielded room (MSR). We have assembled this system and characerized its performances with a mapping robot. The apparatus is able to generate a 1 $\mu$ T vertical field with a relative root mean square deviation $\sigma$ ($B_z$)/$B_z$ = 3 $\times$ $10^{-5}$ over the volume of interest, a cylinder of radius 40 cm and height 30 cm. This level of uniformity overcomes the n2EDM requirements, allowing a measurement of the neutron Electric Dipole Moment with a sensitivity better than 1 $\times$ $10^{-27}$ ecm., Comment: 20 pages, 13 figures

Deep Reinforcement Learning (DRL) in simulation often results in brittle and unrealistic learning outcomes. To push the agent towards more desirable solutions, prior information can be injected in the learning process through, for instance, reward shaping, expert data, or motion primitives. We propose an additional inductive bias for robot learning: latent actions learned from expert demonstration as priors in the action space. We show that these action priors can be learned from only a single open-loop gait cycle using a simple autoencoder. Using these latent action priors combined with established style rewards for imitation in DRL achieves above expert demonstration level of performance and leads to more desirable gaits. Further, action priors substantially improve the performance on transfer tasks, even leading to gait transitions for higher target speeds. Videos and code are available at https://sites.google.com/view/latent-action-priors., Comment: Submitted to ICRA 2025

This paper explores the application of the Condorcet Jury theorem to the domain of sentiment analysis, specifically examining the performance of various large language models (LLMs) compared to simpler natural language processing (NLP) models. The theorem posits that a majority vote classifier should enhance predictive accuracy, provided that individual classifiers' decisions are independent. Our empirical study tests this theoretical framework by implementing a majority vote mechanism across different models, including advanced LLMs such as ChatGPT 4. Contrary to expectations, the results reveal only marginal improvements in performance when incorporating larger models, suggesting a lack of independence among them. This finding aligns with the hypothesis that despite their complexity, LLMs do not significantly outperform simpler models in reasoning tasks within sentiment analysis, showing the practical limits of model independence in the context of advanced NLP tasks.

In this paper, we demonstrate that non-generative, small-sized models such as FinBERT and FinDRoBERTa, when fine-tuned, can outperform GPT-3.5 and GPT-4 models in zero-shot learning settings in sentiment analysis for financial news. These fine-tuned models show comparable results to GPT-3.5 when it is fine-tuned on the task of determining market sentiment from daily financial news summaries sourced from Bloomberg. To fine-tune and compare these models, we created a novel database, which assigns a market score to each piece of news without human interpretation bias, systematically identifying the mentioned companies and analyzing whether their stocks have gone up, down, or remained neutral. Furthermore, the paper shows that the assumptions of Condorcet's Jury Theorem do not hold suggesting that fine-tuned small models are not independent of the fine-tuned GPT models, indicating behavioural similarities. Lastly, the resulted fine-tuned models are made publicly available on HuggingFace, providing a resource for further research in financial sentiment analysis and text classification.

Deep learning models for plant species identification rely on large annotated datasets. The PlantNet system enables global data collection by allowing users to upload and annotate plant observations, leading to noisy labels due to diverse user skills. Achieving consensus is crucial for training, but the vast scale of collected data makes traditional label aggregation strategies challenging. Existing methods either retain all observations, resulting in noisy training data or selectively keep those with sufficient votes, discarding valuable information. Additionally, as many species are rarely observed, user expertise can not be evaluated as an inter-user agreement: otherwise, botanical experts would have a lower weight in the AI training step than the average user. Our proposed label aggregation strategy aims to cooperatively train plant identification AI models. This strategy estimates user expertise as a trust score per user based on their ability to identify plant species from crowdsourced data. The trust score is recursively estimated from correctly identified species given the current estimated labels. This interpretable score exploits botanical experts' knowledge and the heterogeneity of users. Subsequently, our strategy removes unreliable observations but retains those with limited trusted annotations, unlike other approaches. We evaluate PlantNet's strategy on a released large subset of the PlantNet database focused on European flora, comprising over 6M observations and 800K users. We demonstrate that estimating users' skills based on the diversity of their expertise enhances labeling performance. Our findings emphasize the synergy of human annotation and data filtering in improving AI performance for a refined dataset. We explore incorporating AI-based votes alongside human input. This can further enhance human-AI interactions to detect unreliable observations.

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition characterized by impairments in social interaction, communication, as well as restrained or stereotyped behaviors. The inherent heterogeneity within the autism spectrum poses challenges for developing effective pharmacological treatments targeting core features. Successful clinical trials require the identification of robust markers to enable patient stratification. In this study, we explored molecular markers within the oxytocin and immediate early gene families across five interconnected brain structures of the social circuit in four distinct ASD mouse models, each exhibiting unique behavioral features along the autism spectrum. While dysregulations in the oxytocin family were model-specific, immediate early genes displayed widespread alterations, reflecting global changes in social plasticity. Through integrative analysis, we identified Egr1, Foxp1, Homer1a, Oxt and Oxtr as five robust and discriminant molecular markers facilitating successful stratification of the four models. Importantly, our stratification demonstrated predictive values when challenged with a fifth mouse model or identifying subgroups of mice potentially responsive to oxytocin treatment. Beyond providing insights into oxytocin and immediate early gene mRNA dynamics, this proof-of-concept study represents a significant step toward potential stratification of individuals with ASD. The implications extend to enhancing the success of clinical trials and guiding personalized medicine for distinct subgroups of individuals with autism.

This paper introduces a new risk-on risk-off strategy for the stock market, which combines a financial stress indicator with a sentiment analysis done by ChatGPT reading and interpreting Bloomberg daily market summaries. Forecasts of market stress derived from volatility and credit spreads are enhanced when combined with the financial news sentiment derived from GPT-4. As a result, the strategy shows improved performance, evidenced by higher Sharpe ratio and reduced maximum drawdowns. The improved performance is consistent across the NASDAQ, the S&P 500 and the six major equity markets, indicating that the method generalises across equities markets.

We propose a novel approach using Multiplex-Coherent Anti-Stokes Raman Scattering (M-CARS) for la-bel-free discriminations in biomedical tissues. The strategy is based on the evaluation of the contrast be-tween resonant and nonresonant contributions in a M-CARS hyperspectral dataset, and tested to identify and differentiate thin actin filaments from thick myosin filaments in muscle tissue without any labeling. First step consists in ensuring knowledge of the spatial regions containing thick myosin filaments thanks to its endogenous second harmonic signal, deducing expected location for thin actin filaments between myosin filaments. The ratio of resonant and nonresonant contributions for each pixel of the hyperspectral image allows then to discriminate actin from myosin filaments, whose localization is in accordance with the SHG probing. This qualitative imaging represents a proof of principle for highlighting and discriminat-ing purposes in biological microscopy, thanks to the difference in the nonlinear properties of the related proteins. This paves the way for considering label-free imaging through a competition between two third-order nonlinear signatures., Comment: 12 pages, 5 figures

We used a dataset of daily Bloomberg Financial Market Summaries from 2010 to 2023, reposted on large financial media, to determine how global news headlines may affect stock market movements using ChatGPT and a two-stage prompt approach. We document a statistically significant positive correlation between the sentiment score and future equity market returns over short to medium term, which reverts to a negative correlation over longer horizons. Validation of this correlation pattern across multiple equity markets indicates its robustness across equity regions and resilience to non-linearity, evidenced by comparison of Pearson and Spearman correlations. Finally, we provide an estimate of the optimal horizon that strikes a balance between reactivity to new information and correlation.

In multi-photon microscopy (MPM), a recent in-vivo fluorescence microscopy system, the task of image restoration can be decomposed into two interlinked inverse problems: firstly, the characterization of the Point Spread Function (PSF) and subsequently, the deconvolution (i.e., deblurring) to remove the PSF effect, and reduce noise. The acquired MPM image quality is critically affected by PSF blurring and intense noise. The PSF in MPM is highly spread in 3D and is not well characterized, presenting high variability with respect to the observed objects. This makes the restoration of MPM images challenging. Common PSF estimation methods in fluorescence microscopy, including MPM, involve capturing images of sub-resolution beads, followed by quantifying the resulting ellipsoidal 3D spot. In this work, we revisit this approach, coping with its inherent limitations in terms of accuracy and practicality. We estimate the PSF from the observation of relatively large beads (approximately 1$\mu$m in diameter). This goes through the formulation and resolution of an original non-convex minimization problem, for which we propose a proximal alternating method along with convergence guarantees. Following the PSF estimation step, we then introduce an innovative strategy to deal with the high level multiplicative noise degrading the acquisitions. We rely on a heteroscedastic noise model for which we estimate the parameters. We then solve a constrained optimization problem to restore the image, accounting for the estimated PSF and noise, while allowing a minimal hyper-parameter tuning. Theoretical guarantees are given for the restoration algorithm. These algorithmic contributions lead to an end-to-end pipeline for 3D image restoration in MPM, that we share as a publicly available Python software. We demonstrate its effectiveness through several experiments on both simulated and real data.

Autism spectrum disorders are complex, polygenic and heterogenous neurodevelopmental conditions, imposing a substantial economic burden. Genetics are influenced by the environment, specifically the social experience during the critical neurodevelopmental period. Despite efficacy of early behavior interventions targeted specific behaviors in some autistic children, there is no sustainable treatment for the two core symptoms: deficits in social interaction and communication, and stereotyped or restrained behaviors or interests. In this study, we investigated the impact of the social environment on both wild-type (WT) and Shank3 knockout (KO) mice, a mouse model that reproduces core autism-like symptoms. Our findings revealed that WT mice raised in an enriched social environment maintained social interest towards new conspecifics across multiple trials. Additionally, we observed that 2 hours or chronic social isolation induced social deficits or enhanced social interaction and olfactory neuron responses in WT animals, respectively. Notably, chronic social isolation restored both social novelty and olfactory deficits, and normalized self-grooming behavior in Shank3 KO mice. These results novel insights for the implementation of behavioral intervention and inclusive classrooms programs for children with ASD.

High-precision searches for an electric dipole moment of the neutron (nEDM) require stable and uniform magnetic field environments. We present the recent achievements of degaussing and equilibrating the magnetically shielded room (MSR) for the n2EDM experiment at the Paul Scherrer Institute. We present the final degaussing configuration that will be used for n2EDM after numerous studies. The optimized procedure results in a residual magnetic field that has been reduced by a factor of two. The ultra-low field is achieved with the full magnetic-field-coil system, and a large vacuum vessel installed, both in the MSR. In the inner volume of ~1.4 m^3, the field is now more uniform and below 300 pT. In addition, the procedure is faster and dissipates less heat into the magnetic environment, which in turn, reduces its thermal relaxation time from 12 h down to ~1.5 h.

We explain the methodology we developed for improving the interactions accomplished by an embedded conversational agent, drawing from Conversation Analytic sequential and multimodal analysis. The use case is a Pepper robot that is expected to inform and orient users in a library. In order to propose and learn better interactive schema, we are creating a corpus of naturally-occurring interactions that will be made available to the community. To do so, we propose an annotation practice based on some theoretical underpinnings about the use of language and multimodal resources in human-robot interaction. CCS CONCEPTS $\bullet$ Computing methodologies $\rightarrow$ Discourse, dialogue and pragmatics; $\bullet$ Human-centered computing $\rightarrow$ Text input; HCI theory, concepts and models; Field studies., Comment: Peer-reviewed workshop paper accepted for the ''Human-Robot Conversational Interaction'' workshop that took place at the ''ACM/IEEE International Conference on Human-Robot Interaction'' 2023 Conference in Stockholm, Sweden

We present a novel Active Magnetic Shield (AMS), designed and implemented for the n2EDM experiment at the Paul Scherrer Institute. The experiment will perform a high-sensitivity search for the electric dipole moment of the neutron. Magnetic-field stability and control is of key importance for n2EDM. A large, cubic, 5m side length, magnetically shielded room (MSR) provides a passive, quasi-static shielding-factor of about 10^5 for its inner sensitive volume. The AMS consists of a system of eight complex, feedback-controlled compensation coils constructed on an irregular grid spanned on a volume of less than 1000m^3 around the MSR. The AMS is designed to provide a stable and uniform magnetic-field environment around the MSR, while being reasonably compact. The system can compensate static and variable magnetic fields up to +-50muT (homogeneous components) and +-5muT (first-order gradients), suppressing them to a few muT in the sub-Hertz frequency range. The presented design concept and implementation of the AMS fulfills the requirements of the n2EDM experiment and can be useful for other applications, where magnetically silent environments are important and spatial constraints inhibit simpler geometrical solutions.

Automated extraction of quantitative parameters from Cardiac Magnetic Resonance Images (CMRI) is crucial for the management of patients with myocardial infarct. This work proposes a post-processing procedure to jointly analyze Cine and Delayed-Enhanced (DE) acquisitions in order to provide an automatic quantification of myocardial contraction and enhancement parameters and a study of their relationship. For that purpose, the following processes are performed: 1) DE/Cine temporal synchronization and 3D scan alignment, 2) 3D DE/Cine rigid registration in a region about the heart, 3) segmentation of the myocardium on Cine MRI and superimposition of the epicardial and endocardial contours on the DE images, 4) quantification of the Myocardial Infarct Extent (MIE), 5) study of the regional contractile function using a new index, the Amplitude to Time Ratio (ATR). The whole procedure was applied to 10 patients with clinically proven myocardial infarction. The comparison between the MIE and the visually assessed regional function scores demonstrated that the MIE is highly related to the severity of the wall motion abnormality. In addition, it was shown that the newly developed regional myocardial contraction parameter (ATR) decreases significantly in delayed enhanced regions. This largely automated approach enables a combined study of regional MIE and left ventricular function.

We report on a search for a new, short-range, spin-dependent interaction using a modified version of the experimental apparatus used to measure the permanent neutron electric dipole moment at the Paul Scherrer Institute. This interaction, which could be mediated by axion-like particles, concerned the unpolarized nucleons (protons and neutrons) near the material surfaces of the apparatus and polarized ultracold neutrons stored in vacuum. The dominant systematic uncertainty resulting from magnetic-field gradients was controlled to an unprecedented level of approximately 4 pT/cm using an array of optically-pumped cesium vapor magnetometers and magnetic-field maps independently recorded using a dedicated measurement device. No signature of a theoretically predicted new interaction was found, and we set a new limit on the product of the scalar and the pseudoscalar couplings $g_sg_p\lambda^2 < 8.3 \times 10^{-28}\,\text{m}^2$ (95% C.L.) in a range of $5\,\mu\text{m} < \lambda < 25\,\text{mm}$ for the monopole-dipole interaction. This new result confirms and improves our previous limit by a factor of 2.7 and provides the current tightest limit obtained with free neutrons.

We demonstrate how a narrowband pump and a broadband spectrum can be spatially multiplexed by selective coupling them in two distinct modes of a few-mode microstructure fiber. The first mode carries most of the input pump energy, and experiences spectral broadening. Whereas the second mode preserves the narrow bandwidth of the remaining part of the pump. Bimodal propagation, with a power unbalance strongly in favor of the fundamental mode, is naturally obtained by maximizing coupling into the fundamental mode of the fiber. At the fiber output, the nearly monochromatic beam and the supercontinuum carried by the two different modes are combined by a microscope objective, and used as a pump and a Stokes wave for self-referenced multiplex coherent anti-Stokes Raman scattering micro-spectroscopy. The spectral resolution, the signal-to-noise-ratio, and the possible amplification of the remaining pump beam are discussed., Comment: 10 pages, 9 figures

Models that postulate the existence of hidden sectors address contemporary questions, such as the source of baryogenesis and the nature of dark matter. Among the possible mixing processes, neutron-to-hidden-neutron oscillations have been repeatedly tested with ultra-cold neutron storage and passing-through-wall experiments in the range of small ($\delta m<2$ peV) and large mass splitting ($\delta m>10$ neV), respectively. In this work, we present a new constraint in the oscillation parameter space derived from neutron disappearance in ultra-cold neutron beam experiments. The overall limit, which covers the intermediate mass-splitting range, is given by $\tau_{nn'}> 1$ s for $|\delta m| \in [2,69]$ peV (95\% C.L.).

Ultra-low-mass axions are a viable dark matter candidate and may form a coherently oscillating classical field. Nuclear spins in experiments on Earth might couple to this oscillating axion dark-matter field, when propagating on Earth's trajectory through our Galaxy. This spin coupling resembles an oscillating pseudo-magnetic field which modulates the spin precession of nuclear spins. Here we report on the null result of a demonstration experiment searching for a frequency modulation of the free spin-precession signal of \magHg in a \SI{1}{\micro\tesla} magnetic field. Our search covers the axion mass range $10^{-16}~\textrm{eV} \lesssim m_a \lesssim 10^{-13}~\textrm{eV}$ and achieves a peak sensitivity to the axion-nucleon coupling of $g_{aNN} \approx 3.5 \times 10^{-6}~\textrm{GeV}^{-1}$., Comment: 18 pages, 4 images, submitted to SciPost Physics

The ability to accurately detect onset of dementia is important in the treatment of the disease. Clinically, the diagnosis of Alzheimer Disease (AD) and Mild Cognitive Impairment (MCI) patients are based on an integrated assessment of psychological tests and brain imaging such as positron emission tomography (PET) and anatomical magnetic resonance imaging (MRI). In this work using two different datasets, we propose a behavior score-embedded encoder network (BSEN) that integrates regularly adminstrated psychological tests information into the encoding procedure of representing subject's restingstate fMRI data for automatic classification tasks. BSEN is based on a 3D convolutional autoencoder structure with contrastive loss jointly optimized using behavior scores from MiniMental State Examination (MMSE) and Clinical Dementia Rating (CDR). Our proposed classification framework of using BSEN achieved an overall recognition accuracy of 59.44% (3-class classification: AD, MCI and Healthy Control), and we further extracted the most discriminative regions between healthy control (HC) and AD patients., Comment: 4 pages, 1 figure

Self-supervised visual representation methods are closing the gap with supervised learning performance. These methods rely on maximizing the similarity between embeddings of related synthetic inputs created through data augmentations. This can be seen as a task that encourages embeddings to leave out factors modified by these augmentations, i.e. to be invariant to them. However, this only considers one side of the trade-off in the choice of the augmentations: they need to strongly modify the images to avoid simple solution shortcut learning (e.g. using only color histograms), but on the other hand, augmentations-related information may be lacking in the representations for some downstream tasks (e.g. color is important for birds and flower classification). Few recent works proposed to mitigate the problem of using only an invariance task by exploring some form of equivariance to augmentations. This has been performed by learning additional embeddings space(s), where some augmentation(s) cause embeddings to differ, yet in a non-controlled way. In this work, we introduce EquiMod a generic equivariance module that structures the learned latent space, in the sense that our module learns to predict the displacement in the embedding space caused by the augmentations. We show that applying that module to state-of-the-art invariance models, such as SimCLR and BYOL, increases the performances on CIFAR10 and ImageNet datasets. Moreover, while our model could collapse to a trivial equivariance, i.e. invariance, we observe that it instead automatically learns to keep some augmentations-related information beneficial to the representations.

In supervised learning - for instance in image classification - modern massive datasets are commonly labeled by a crowd of workers. The obtained labels in this crowdsourcing setting are then aggregated for training, generally leveraging a per-worker trust score. Yet, such workers oriented approaches discard the tasks' ambiguity. Ambiguous tasks might fool expert workers, which is often harmful for the learning step. In standard supervised learning settings - with one label per task - the Area Under the Margin (AUM) was tailored to identify mislabeled data. We adapt the AUM to identify ambiguous tasks in crowdsourced learning scenarios, introducing the Weighted Areas Under the Margin (WAUM). The WAUM is an average of AUMs weighted according to task-dependent scores. We show that the WAUM can help discarding ambiguous tasks from the training set, leading to better generalization performance. We report improvements over existing strategies for learning with a crowd, both on simulated settings, and on real datasets such as CIFAR-10H (a crowdsourced dataset with a high number of answered labels),LabelMe and Music (two datasets with few answered votes).

Numerical validation is at the core of machine learning research as it allows to assess the actual impact of new methods, and to confirm the agreement between theory and practice. Yet, the rapid development of the field poses several challenges: researchers are confronted with a profusion of methods to compare, limited transparency and consensus on best practices, as well as tedious re-implementation work. As a result, validation is often very partial, which can lead to wrong conclusions that slow down the progress of research. We propose Benchopt, a collaborative framework to automate, reproduce and publish optimization benchmarks in machine learning across programming languages and hardware architectures. Benchopt simplifies benchmarking for the community by providing an off-the-shelf tool for running, sharing and extending experiments. To demonstrate its broad usability, we showcase benchmarks on three standard learning tasks: $\ell_2$-regularized logistic regression, Lasso, and ResNet18 training for image classification. These benchmarks highlight key practical findings that give a more nuanced view of the state-of-the-art for these problems, showing that for practical evaluation, the devil is in the details. We hope that Benchopt will foster collaborative work in the community hence improving the reproducibility of research findings., Comment: Accepted in proceedings of NeurIPS 22; Benchopt library documentation is available at https://benchopt.github.io/