Search

Designers often engage with video to gain rich, temporal insights about the context of users, collaboratively analyzing it to gather ideas, challenge assumptions, and foster empathy. To capture the full visual context of users and their situations, designers are adopting 360$^\circ$ video, providing richer, more multi-layered insights. Unfortunately, the spherical nature of 360$^\circ$ video means designers cannot create tangible video artifacts such as storyboards for collaborative analysis. To overcome this limitation, we created Tangi, a web-based tool that converts 360$^\circ$ images into tangible 360$^\circ$ video artifacts, that enable designers to embody and share their insights. Our evaluation with nine experienced designers demonstrates that the artifacts Tangi creates enable tangible interactions found in collaborative workshops and introduce two new capabilities: spatial orientation within 360$^\circ$ environments and linking specific details to the broader 360$^\circ$ context. Since Tangi is an open-source tool, designers can immediately leverage 360$^\circ$ video in collaborative workshops.

Sterile neutrinos are well-motivated beyond the Standard Model (BSM) particles. The Standard Model Effective Field Theory (SMEFT) augmented with these new fields is known as the $\nu$SMEFT. We present the first code for solving the renormalization group equations (RGEs) of the $\nu$SMEFT in an automated way. For this purpose, we have implemented the $\nu$SMEFT as a new effective field theory (EFT) in the Wilson coefficient exchange format WCxf. Furthermore, we included anomalous dimensions depending on the gauge couplings and Yukawas in the python package wilson. This novel version of wilson allows a consistent inclusion of $\nu$SMEFT renormalization group (RG) running effects above the electroweak (EW) scale in phenomenological studies involving sterile neutrinos. Moreover, this new release allows us to study EW, strong, and Yukawa running effects separately within the SMEFT., Comment: 14 pages

Predicting temporal progress from visual trajectories is important for intelligent robots that can learn, adapt, and improve. However, learning such progress estimator, or temporal value function, across different tasks and domains requires both a large amount of diverse data and methods which can scale and generalize. To address these challenges, we present Generative Value Learning (\GVL), a universal value function estimator that leverages the world knowledge embedded in vision-language models (VLMs) to predict task progress. Naively asking a VLM to predict values for a video sequence performs poorly due to the strong temporal correlation between successive frames. Instead, GVL poses value estimation as a temporal ordering problem over shuffled video frames; this seemingly more challenging task encourages VLMs to more fully exploit their underlying semantic and temporal grounding capabilities to differentiate frames based on their perceived task progress, consequently producing significantly better value predictions. Without any robot or task specific training, GVL can in-context zero-shot and few-shot predict effective values for more than 300 distinct real-world tasks across diverse robot platforms, including challenging bimanual manipulation tasks. Furthermore, we demonstrate that GVL permits flexible multi-modal in-context learning via examples from heterogeneous tasks and embodiments, such as human videos. The generality of GVL enables various downstream applications pertinent to visuomotor policy learning, including dataset filtering, success detection, and advantage-weighted regression -- all without any model training or finetuning., Comment: Project website and demo: https://generative-value-learning.github.io/

Automated Machine Learning (AutoML) approaches encompass traditional methods that optimize fixed pipelines for model selection and ensembling, as well as newer LLM-based frameworks that autonomously build pipelines. While LLM-based agents have shown promise in automating machine learning tasks, they often generate low-diversity and suboptimal code, even after multiple iterations. To overcome these limitations, we introduce Tree-Search Enhanced LLM Agents (SELA), an innovative agent-based system that leverages Monte Carlo Tree Search (MCTS) to optimize the AutoML process. By representing pipeline configurations as trees, our framework enables agents to conduct experiments intelligently and iteratively refine their strategies, facilitating a more effective exploration of the machine learning solution space. This novel approach allows SELA to discover optimal pathways based on experimental feedback, improving the overall quality of the solutions. In an extensive evaluation across 20 machine learning datasets, we compare the performance of traditional and agent-based AutoML methods, demonstrating that SELA achieves a win rate of 65% to 80% against each baseline across all datasets. These results underscore the significant potential of agent-based strategies in AutoML, offering a fresh perspective on tackling complex machine learning challenges., Comment: The code is available at https://github.com/geekan/MetaGPT

Coding tasks have been valuable for evaluating Large Language Models (LLMs), as they demand the comprehension of high-level instructions, complex reasoning, and the implementation of functional programs -- core capabilities for advancing Artificial General Intelligence. Despite the progress in Large Multimodal Models (LMMs), which extend LLMs with visual perception and understanding capabilities, there remains a notable lack of coding benchmarks that rigorously assess these models, particularly in tasks that emphasize visual reasoning. To address this gap, we introduce HumanEval-V, a novel and lightweight benchmark specifically designed to evaluate LMMs' visual understanding and reasoning capabilities through code generation. HumanEval-V includes 108 carefully crafted, entry-level Python coding tasks derived from platforms like CodeForces and Stack Overflow. Each task is adapted by modifying the context and algorithmic patterns of the original problems, with visual elements redrawn to ensure distinction from the source, preventing potential data leakage. LMMs are required to complete the code solution based on the provided visual context and a predefined Python function signature outlining the task requirements. Every task is equipped with meticulously handcrafted test cases to ensure a thorough and reliable evaluation of model-generated solutions. We evaluate 19 state-of-the-art LMMs using HumanEval-V, uncovering significant challenges. Proprietary models like GPT-4o achieve only 13% pass@1 and 36.4% pass@10, while open-weight models with 70B parameters score below 4% pass@1. Ablation studies further reveal the limitations of current LMMs in vision reasoning and coding capabilities. These results underscore key areas for future research to enhance LMMs' capabilities. We have open-sourced our code and benchmark at https://github.com/HumanEval-V/HumanEval-V-Benchmark., Comment: homepage https://humaneval-v.github.io/

We investigate the dynamics of short-range interacting Bose gases with varying degrees of diluteness and interaction strength. By applying a combined mean-field and semiclassical space-time rescaling to the dynamics in both the Gross--Pitaevskii and hard-core limit regimes, we prove that the local one-particle mass, momentum, and energy densities of the many-body system can be quantitatively approximated by solutions to the compressible Euler system in the strong sense, up to the first blow-up time of the fluid description, as the number of particles tends to infinity. In the hard-core limit regime, two novel results are presented. First, we rigorously prove, for the first time, that the internal energy of the fluid takes the form $4\pi \mathfrak{c}_{0}\rho^{2}$ (equivalently, pressure $P=2\pi \mathfrak{c}_{0}\rho^{2}$), arising solely from the kinetic energy density of the many-body system, rather than the interaction energy density, marking a fundamental difference from the Gross--Pitaevskii and other mean-field regimes. Second, the newly discovered coupling constant $\mathfrak{c}_{0}$ is the electrostatic capacity of the interaction potential, corresponding to the scattering length of the hard-core potential. Furthermore, in other limiting regimes, including those beyond the Gross--Pitaevskii regime, we find that the limiting equation is described by an eikonal system, offering a rigorous first-principle justification for using the ``geometric optics approximation'' to describe the dynamics of ultracold Bose gases.

Hot, dense phases of Dirac fermions - predicted to resemble relativistic plasma - are uniquely accessible through photoexcitation of pristine, charge neutral graphene. We demonstrate a sensitive temperature probe of the photoexcited Dirac state, called interlayer optoelectronic thermometry, which measures out-of-plane transport of hot carriers in high-mobility, neutral graphene encapsulated within graphene-hBN-graphene heterostructures. At a critical intermediate sample temperature T = 50 K, the electronic temperature Te is quenched, exhibiting an intrinsic cooling rate that exceeds 10^14 Kelvin/s within the first picosecond after photoexcitation. Quenching is further enhanced by applying in-plane voltages within the stack-engineered heterostructure. Extreme sensitivity of Te to sample temperature and applied voltages reveals anomalously efficient hot-carrier quenching, which we identify as an essential feature of the strongly interacting hot Dirac excited state.

Backdoor attacks, in which a model behaves maliciously when given an attacker-specified trigger, pose a major security risk for practitioners who depend on publicly released language models. Backdoor detection methods aim to detect whether a released model contains a backdoor, so that practitioners can avoid such vulnerabilities. While existing backdoor detection methods have high accuracy in detecting backdoored models on standard benchmarks, it is unclear whether they can robustly identify backdoors in the wild. In this paper, we examine the robustness of backdoor detectors by manipulating different factors during backdoor planting. We find that the success of existing methods highly depends on how intensely the model is trained on poisoned data during backdoor planting. Specifically, backdoors planted with either more aggressive or more conservative training are significantly more difficult to detect than the default ones. Our results highlight a lack of robustness of existing backdoor detectors and the limitations in current benchmark construction.

Recent advancements in Distributional Reinforcement Learning (DRL) for modeling loss distributions have shown promise in developing hedging strategies in derivatives markets. A common approach in DRL involves learning the quantiles of loss distributions at specified levels using Quantile Regression (QR). This method is particularly effective in option hedging due to its direct quantile-based risk assessment, such as Value at Risk (VaR) and Conditional Value at Risk (CVaR). However, these risk measures depend on the accurate estimation of extreme quantiles in the loss distribution's tail, which can be imprecise in QR-based DRL due to the rarity and extremity of tail data, as highlighted in the literature. To address this issue, we propose EXtreme DRL (EX-DRL), which enhances extreme quantile prediction by modeling the tail of the loss distribution with a Generalized Pareto Distribution (GPD). This method introduces supplementary data to mitigate the scarcity of extreme quantile observations, thereby improving estimation accuracy through QR. Comprehensive experiments on gamma hedging options demonstrate that EX-DRL improves existing QR-based models by providing more precise estimates of extreme quantiles, thereby improving the computation and reliability of risk metrics for complex financial risk management., Comment: 14 pages

Two categories of four-fermion SMEFT operators are semileptonic (two quarks and two leptons) and hadronic (four quarks). At tree level, an operator of a given category contributes only to processes of the same category. However, when the SMEFT Hamiltonian is evolved down from the new-physics scale to low energies using the renormalization-group equations (RGEs), due to operator mixing this same SMEFT operator can generate operators of the other category at one loop. Thus, to search for a SMEFT explanation of a low-energy anomaly, or combination of anomalies, one must: (i) identify the candidate semileptonic and hadronic SMEFT operators, (ii) run them down to low energy with the RGEs, (iii) generate the required low-energy operators with the correct Wilson coefficients, and (iv) check that all other constraints are satisfied. In this paper, we illustrate this method by finding all SMEFT operators that, by themselves, provide a combined explanation of the (semileptonic) $\bar b \to \bar s \ell^+ \ell^-$ anomalies and the (hadronic) $B \to K \pi$ puzzle., Comment: 29 pages, no figures

Heavy fermion quantum criticality is an extremely rich domain of research which represents a framework to understand strange metals as a consequence of a Kondo breakdown transition. Here we provide an experimental realization of such systems in terms of organic radicals on a metallic surface. The ground state of organic radicals is a Kramer's doublet that can be modeled by a spin 1/2 degree of freedom. Using on-surface synthesis and scanning tunneling microscopy (STM) tip manipulation, one can controllably engineer and characterize chains of organic radicals on a Au(111) surface. The spatial-resolved differential conductance reveals site-dependent low-energy excitations, which support the picture of emergent many-body Kondo physics. Using quantum Monte Carlo simulations, we show that a Kondo lattice model of spin chains on a metallic surface reproduces accurately the experimental results. This allows us to interpret the experimental results in terms of a heavy fermion metal, below the coherence temperature. We foresee that the tunability of these systems will pave the way to realize quantum simulators of heavy fermion criticality.

Background: Low vitamin D intake and high prevalence of vitamin D deficiency (serum 25-hydroxyvitamin D concentration < 50 nmol/L) among Aboriginal and Torres Strait Islander peoples highlight a need for public health strategies to improve vitamin D status. As few foods contain naturally occurring vitamin D, fortification strategies may be needed to improve vitamin D intake and status among Aboriginal and Torres Strait Islander peoples. Objective: We aimed to model vitamin D food fortification scenarios among Aboriginal and Torres Strait Islander peoples. Methods: We used nationally representative food consumption data (n=4,109) and vitamin D food composition data to model four food fortification scenarios. The modelling for Scenario 1 included foods and maximum vitamin D concentrations permitted for fortification in Australia: i) dairy products and alternatives, ii) butter/margarine/oil spreads, iii) formulated beverages, and iv) selected ready-to-eat breakfast cereals. The modelling for Scenarios 2a-c included some vitamin D concentrations higher than permitted in Australia; Scenario 2c included bread, which is not permitted for vitamin D fortification in Australia. Scenario 2a: i) dairy products and alternatives, ii) butter/margarine/oil spreads, iii) formulated beverages. Scenario 2b: as per Scenario 2a plus selected ready-to-eat breakfast cereals. Scenario 2c: as per Scenario 2b plus bread. Results: Vitamin D fortification of a range of staple foods could potentially increase vitamin D intake among Aboriginal and Torres Strait Islander peoples by ~ 3-6 {\mu}g/day. Scenario 2c showed the highest potential median vitamin D intake increase to ~ 8 {\mu}g/day. Across all modelled scenarios, none of the participants had vitamin D intake above the Australian upper level of intake of 80 {\mu}g/day.

Accurately predicting carrier mobility in strongly anharmonic solids necessitates a precise characterization of lattice dyndamics as a function of temperature. We achieve consistency with experimental electron mobility data for bulk KTaO$_3$ and SrTiO$_3$ above 150 K by refining the Boltzmann transport equations. This refinement includes incorporating temperature-dependent anharmonic phonon eigenfrequencies and eigenmodes into the electron-phonon interaction tensor, while maintaining the derivatives of the Kohn-Sham potential as computed in density functional perturbation theory. Using efficient machine-learned force fields and the stochastic self-consistent harmonic approximation, we accurately compute the dynamical matrices. At room temperature, the calculated mobility for SrTiO$_3$ exceeds experimental values by an order of magnitude, whereas the overestimation for KTaO$_3$ is way less pronounced. This discrepancy is explained through the more significant electron mass renormalization near the conduction band bottom due to anharmonic electron-phonon coupling and the presence of local disorder in SrTiO$_3$.

Over recent decades, neuroimaging tools, particularly electroencephalography (EEG), have revolutionized our understanding of the brain and its functions. EEG is extensively used in traditional brain-computer interface (BCI) systems due to its low cost, non-invasiveness, and high temporal resolution. This makes it invaluable for identifying different brain states relevant to both medical and non-medical applications. Although this practice is widely recognized, current methods are mainly confined to lab or clinical environments because they rely on data from multiple EEG electrodes covering the entire head. Nonetheless, a significant advancement for these applications would be their adaptation for "real-world" use, using portable devices with a single-channel. In this study, we tackle this challenge through two distinct strategies: the first approach involves training models with data from multiple channels and then testing new trials on data from a single channel individually. The second method focuses on training with data from a single channel and then testing the performances of the models on data from all the other channels individually. To efficiently classify cognitive tasks from EEG data, we propose Convolutional Neural Networks (CNNs) with only a few parameters and fast learnable spectral-temporal features. We demonstrated the feasibility of these approaches on EEG data recorded during mental arithmetic and motor imagery tasks from three datasets. We achieved the highest accuracies of 100%, 91.55% and 73.45% in binary and 3-class classification on specific channels across three datasets. This study can contribute to the development of single-channel BCI and provides a robust EEG biomarker for brain states classification., Comment: Accepted in the 16th International Conference on Human System Interaction 2024, Paris, France

The increased ubiquity of 360{\deg} video presents a unique opportunity for designers to deeply engage with the world of users by capturing the complete visual context. However, the opportunities and challenges 360{\deg} video introduces for video design ethnography is unclear. This study investigates this gap through 16 workshops in which experienced designers engaged with 360{\deg} video. Our analysis shows that while 360{\deg} video enhances designers' ability to explore and understand user contexts, it also complicates the process of sharing insights. To address this challenge, we present two opportunities to support the use of 360{\deg} video by designers - the creation of designerly 360{\deg} video annotation tools, and 360{\deg} ``screenshots'' - in order to enable designers to leverage the complete context of 360{\deg} video for user research.

Background/Objective: Vitamin D deficiency (serum 25-hydroxyvitamin D [25(OH)D] concentration <50 nmol/L) is prevalent among Aboriginal and Torres Strait Islander peoples in Australia. Alternative to sun exposure (the primary source of vitamin D), vitamin D can also be obtained from food (e.g., fish, eggs, and meat) and supplements. However, vitamin D intake among Aboriginal and Torres Strait Islander peoples is currently unknown. We aimed to provide the first quantification of vitamin D intake using nationally representative data from Aboriginal and Torres Strait Islander peoples. Methods: We used food consumption data collected in the 2012-2013 National Aboriginal and Torres Strait Islander Nutrition and Physical Activity Survey (n = 4,109) and vitamin D food composition data to quantify mean absolute vitamin D intake by sex, age group, and remoteness of location. Differences in mean vitamin D intake between sexes and between remoteness of location were assessed using the 95% confidence interval (95% CI). Results: The mean (standard deviation (SD)) vitamin D intake among Aboriginal and Torres Strait Islander peoples was 2.9 (3.0) {\mu}g/day. Males had a statistically significantly higher mean (SD) [95% CI] vitamin D intake (3.2 (3.1) [3.0-3.4] {\mu}g/day) than females (2.6 (2.7) [2.4-2.7] {\mu}g/day). There were no statistically significant differences between mean (SD) [95% CI] vitamin D intake in non-remote (2.9 (2.2) [2.7-3.1] {\mu}g/day) and remote areas (2.8 (4.8) [2.6-3.0] {\mu}g/day). Conclusions: Vitamin D intake among Aboriginal and Torres Strait Islander peoples is low. Food-based public health strategies could be developed to promote higher vitamin D intake among this population.

The increasing use of Large Language Models (LLMs) in software development has garnered significant attention from researchers assessing the quality of the code they generate. However, much of the research focuses on controlled datasets such as HumanEval, which fail to adequately represent how developers actually utilize LLMs' code generation capabilities or clarify the characteristics of LLM-generated code in real-world development scenarios. To bridge this gap, our study investigates the characteristics of LLM-generated code and its corresponding projects hosted on GitHub. Our findings reveal several key insights: (1) ChatGPT and Copilot are the most frequently utilized for generating code on GitHub. In contrast, there is very little code generated by other LLMs on GitHub. (2) Projects containing ChatGPT/Copilot-generated code are often small and less known, led by individuals or small teams. Despite this, most projects are continuously evolving and improving. (3) ChatGPT/Copilot is mainly utilized for generating Python, Java, and TypeScript scripts for data processing and transformation. C/C++ and JavaScript code generation focuses on algorithm and data structure implementation and user interface code. Most ChatGPT/Copilot-generated code snippets are relatively short and exhibit low complexity. (4) Compared to human-written code, ChatGPT/Copilot-generated code exists in a small proportion of projects and generally undergoes fewer modifications. Additionally, modifications due to bugs are even fewer, ranging from just 3% to 8% across different languages. (5) Most comments on ChatGPT/Copilot-generated code lack detailed information, often only stating the code's origin without mentioning prompts, human modifications, or testing status. Based on these findings, we discuss the implications for researchers and practitioners.

Code Smell Detection (CSD) plays a crucial role in improving software quality and maintainability. And Deep Learning (DL) techniques have emerged as a promising approach for CSD due to their superior performance. However, the effectiveness of DL-based CSD methods heavily relies on the quality of the training data. Despite its importance, little attention has been paid to analyzing the data preparation process. This systematic literature review analyzes the data preparation techniques used in DL-based CSD methods. We identify 36 relevant papers published by December 2023 and provide a thorough analysis of the critical considerations in constructing CSD datasets, including data requirements, collection, labeling, and cleaning. We also summarize seven primary challenges and corresponding solutions in the literature. Finally, we offer actionable recommendations for preparing and accessing high-quality CSD data, emphasizing the importance of data diversity, standardization, and accessibility. This survey provides valuable insights for researchers and practitioners to harness the full potential of DL techniques in CSD.

We introduce MuirBench, a comprehensive benchmark that focuses on robust multi-image understanding capabilities of multimodal LLMs. MuirBench consists of 12 diverse multi-image tasks (e.g., scene understanding, ordering) that involve 10 categories of multi-image relations (e.g., multiview, temporal relations). Comprising 11,264 images and 2,600 multiple-choice questions, MuirBench is created in a pairwise manner, where each standard instance is paired with an unanswerable variant that has minimal semantic differences, in order for a reliable assessment. Evaluated upon 20 recent multi-modal LLMs, our results reveal that even the best-performing models like GPT-4o and Gemini Pro find it challenging to solve MuirBench, achieving 68.0% and 49.3% in accuracy. Open-source multimodal LLMs trained on single images can hardly generalize to multi-image questions, hovering below 33.3% in accuracy. These results highlight the importance of MuirBench in encouraging the community to develop multimodal LLMs that can look beyond a single image, suggesting potential pathways for future improvements., Comment: typos corrected, references added, Project Page: https://muirbench.github.io/

Federated Learning (FL) enables local devices to collaboratively learn a shared predictive model by only periodically sharing model parameters with a central aggregator. However, FL can be disadvantaged by statistical heterogeneity produced by the diversity in each local devices data distribution, which creates different levels of Independent and Identically Distributed (IID) data. Furthermore, this can be more complex when optimising different combinations of FL parameters and choosing optimal aggregation. In this paper, we present an empirical analysis of different FL training parameters and aggregators over various levels of statistical heterogeneity on three datasets. We propose a systematic data partition strategy to simulate different levels of statistical heterogeneity and a metric to measure the level of IID. Additionally, we empirically identify the best FL model and key parameters for datasets of different characteristics. On the basis of these, we present recommended guidelines for FL parameters and aggregators to optimise model performance under different levels of IID and with different datasets, Comment: 27 pages, 17 figures

With the increasing utilization of large language models such as ChatGPT during software development, it has become crucial to verify the quality of code content it generates. Recent studies proposed utilizing ChatGPT as both a developer and tester for multi-agent collaborative software development. The multi-agent collaboration empowers ChatGPT to produce test reports for its generated code, enabling it to self-verify the code content and fix bugs based on these reports. However, these studies did not assess the effectiveness of the generated test reports in validating the code. Therefore, we conduct a comprehensive empirical investigation to evaluate ChatGPT's self-verification capability in code generation, code completion, and program repair. We request ChatGPT to (1) generate correct code and then self-verify its correctness; (2) complete code without vulnerabilities and then self-verify for the presence of vulnerabilities; and (3) repair buggy code and then self-verify whether the bugs are resolved. Our findings on two code generation datasets, one code completion dataset, and two program repair datasets reveal the following observations: (1) ChatGPT often erroneously predicts its generated incorrect code as correct. (2) The self-contradictory hallucinations in ChatGPT's behavior arise. (3) The self-verification capability of ChatGPT can be enhanced by asking the guiding question, which queries whether ChatGPT agrees with assertions about incorrectly generated or repaired code and vulnerabilities in completed code. (4) Using test reports generated by ChatGPT can identify more vulnerabilities in completed code, but the explanations for incorrectly generated code and failed repairs are mostly inaccurate in the test reports. Based on these findings, we provide implications for further research or development using ChatGPT.

Numerous Deep Learning (DL)-based approaches have gained attention in software Log Anomaly Detection (LAD), yet class imbalance in training data remains a challenge, with anomalies often comprising less than 1% of datasets like Thunderbird. Existing DLLAD methods may underperform in severely imbalanced datasets. Although data resampling has proven effective in other software engineering tasks, it has not been explored in LAD. This study aims to fill this gap by providing an in-depth analysis of the impact of diverse data resampling methods on existing DLLAD approaches from two distinct perspectives. Firstly, we assess the performance of these DLLAD approaches across four datasets with different levels of class imbalance, and we explore the impact of resampling ratios of normal to abnormal data on DLLAD approaches. Secondly, we evaluate the effectiveness of the data resampling methods when utilizing optimal resampling ratios of normal to abnormal data. Our findings indicate that oversampling methods generally outperform undersampling and hybrid sampling methods. Data resampling on raw data yields superior results compared to data resampling in the feature space. These improvements are attributed to the increased attention given to important tokens. By exploring the resampling ratio of normal to abnormal data, we suggest generating more data for minority classes through oversampling while removing less data from majority classes through undersampling. In conclusion, our study provides valuable insights into the intricate relationship between data resampling methods and DLLAD. By addressing the challenge of class imbalance, researchers and practitioners can enhance DLLAD performance., Comment: 18 pages, 3 figures

Sterile neutrinos with masses at the $\mathrm{keV}$ scale and mixing to the active neutrinos offer an elegant explanation of the observed dark matter (DM) density. However, the very same mixing inevitably leads to radiative photon emission and the non-observation of such peaked $X$-ray lines rules out this minimal sterile neutrino DM hypothesis. We show that in the context of the Standard Model effective field theory with sterile neutrinos ($\nu$SMEFT), higher dimensional operators can produce sterile neutrino DM in a broad range of parameter space. In particular, $\nu$SMEFT interactions can open the large mixing parameter space due to their destructive interference, through operator mixing or matching, in the $X$-ray emission. We also find that, even in the zero mixing limit, the DM density can always be explained by $\nu$SMEFT operators. The testability of the studied $\nu$SMEFT operators in searches for electric dipole moments, neutrinoless double beta decay, and pion decay measurements is discussed., Comment: 28 pages + 4 pages appendix, 9 figures

Code translation tools (transpilers) are developed for automatic source-to-source translation. Although learning-based transpilers have shown impressive enhancement against rule-based counterparts, owing to their task-specific pre-training on extensive monolingual corpora. Their current performance still remains unsatisfactory for practical deployment, and the associated training resources are also prohibitively expensive. LLMs pre-trained on huge amounts of human-written code/text have shown remarkable performance in many code intelligence tasks due to their powerful generality, even without task-specific training. Thus, LLMs can potentially circumvent the above limitations, but they have not been exhaustively explored yet. This paper investigates diverse LLMs and learning-based transpilers for automated code translation tasks, finding that: although certain LLMs have outperformed current transpilers, they still have some accuracy issues, where most of the failures are induced by a lack of comprehension of source programs, missing clear instructions on I/O types in translation, and ignoring discrepancies between source and target programs. Enlightened by the above findings, we further propose UniTrans, a Unified code Translation framework, applicable to various LLMs, for unleashing their power in this field. Specifically, UniTrans first crafts a series of test cases for target programs with the assistance of source programs. Next, it harnesses the above auto-generated test cases to augment the code translation and then evaluate their correctness via execution. Afterward, UniTrans further (iteratively) repairs incorrectly translated programs prompted by test case execution results. Extensive experiments are conducted on six settings of translation datasets between Python, Java, and C++. Three recent LLMs of diverse sizes are tested with UniTrans, and all achieve substantial improvements., Comment: 23 pages, 7 figures, accepted by FSE'24 (2024 ACM International Conference on the Foundations of Software Engineering)

Layered halide perovskites have emerged as potential alternatives to three-dimensional halide perovskites due to their improved stability and larger material phase space, allowing fine-tuning of structural, electronic, and optical properties. However, their charge carrier mobilities are significantly smaller than that of three-dimensional halide perovskites, which has a considerable impact on their application in optoelectronic devices. Here, we employ state-of-the-art ab initio approaches to unveil the electron-phonon mechanisms responsible for the diminished transport properties of layered halide perovskites. Starting from a prototypical ABX$_{3}$ halide perovskite, we model the case of $n=1$ and $n=2$ layered structures and compare their electronic and transport properties to the three-dimensional reference. The electronic and phononic properties are investigated within density functional theory (DFT) and density functional perturbation theory (DFPT), while transport properties are obtained via the ab initio Boltzmann transport equation. The vibrational modes contributing to charge carrier scattering are investigated and associated with polar-phonon scattering mechanisms arising from the long-range Fr\"ohlich coupling and deformation potential scattering processes. Our investigation reveals that the lower mobilities in layered systems primarily originates from the increased electronic density of states at the vicinity of the band edges, while the electron-phonon coupling strength remains similar. Such increase is caused by the dimensionality reduction and the break in octahedra connectivity along the stacking direction. Our findings provide a fundamental understanding of the electron-phonon coupling mechanisms in layered perovskites and highlight the intrinsic limitations of the charge carrier transport in these materials.

Lead halide perovskite quantum dots (QDs), the latest generation of colloidal QD family, exhibit outstanding optical properties which are now exploited as both classical and quantum light sources. Most of their rather exceptional properties are related to the peculiar exciton fine-structure of band-edge states which can support unique bright triplet excitons. The degeneracy of the bright triplet excitons is lifted with energetic splitting in the order of millielectronvolts, which can be resolved by the photoluminescence (PL) measurements of single QDs at cryogenic temperatures. Each bright exciton fine-structure-state (FSS) exhibits a dominantly linear polarization, in line with several theoretical models based on the sole crystal field, exchange interaction and shape anisotropy. Here, we show that in addition to a high degree of linear polarization, the individual exciton FSS can exhibit a non-negligible degree of circular polarization even without external magnetic fields by investigating the four Stokes parameters of the exciton fine-structure in individual CsPbBr3 QDs through Stokes polarimetric measurements. We observe a degree of circular polarization up to ~38%, which could not be detected by using the conventional polarimetric technique. In addition, we found a consistent transition from left- to right-hand circular polarization within the fine-structure triplet manifold, which was observed in magnetic field dependent experiments. Our optical investigation provides deeper insights into the nature of the exciton fine-structures and thereby drives the yet-incomplete understanding of the unique photophysical properties of this novel class of QDs, potentially opening new scenarios in chiral quantum optics.

A fully generic treatment of electric dipole moments (EDMs) is presented in the CP-violating and flavor-conserving weak effective field theory (WET) with five flavors of quarks and three flavors of leptons. We systematically analyze leading contributions to EDMs originating from QCD and QED renormalization group running between the electroweak scale and low energy scales of about 2 GeV. We include the full one-loop anomalous dimension and a subset of two-loop corrections, as well as threshold corrections at the bottom, charm and $\tau$ masses. This allows us to derive master formulae in the space of generic WET for the neutron and proton EDMs, for EDMs of diamagnetic atoms, and the precession frequencies constrained in molecular EDM experiments, from which bounds on the electron EDM are extracted. In particular, our master formulae capture the contributions of WET CP-violating operators with heavy quark and lepton flavors. As an application, we study EDM constraints on the Yukawa couplings of the Higgs boson, in both the linear and non-linear realizations of electroweak symmetry breaking., Comment: 60 pages, 5 figures, text improved, bounds on electron Yukawas added, part of numerics updated

In this paper, we rewrite the time-dependent Bogoliubov$\unicode{x2013}$de Gennes equation in an appropriate semiclassical form and establish its semiclassical limit to a two-particle kinetic transport equation with an effective mean-field background potential satisfying the one-particle Vlasov equation. Moreover, for some semiclassical regimes, we obtain a higher-order correction to the two-particle kinetic transport equation, capturing a nontrivial two-body interaction effect. The convergence is proven for $C^2$ interaction potentials in terms of a semiclassical optimal transport pseudo-metric. Furthermore, combining our current results with the results of Marcantoni et al. [arXiv:2310.15280], we establish a joint semiclassical and mean-field approximation of the dynamics of a system of spin-$\frac{1}{2}$ Fermions by the Vlasov equation in some negative order Sobolev topology., Comment: 29 pages

Persistent homology naturally addresses the multi-scale topological characteristics of the large-scale structure as a distribution of clusters, loops, and voids. We apply this tool to the dark matter halo catalogs from the Quijote simulations, and build a summary statistic for comparison with the joint power spectrum and bispectrum statistic regarding their information content on cosmological parameters and primordial non-Gaussianity. Through a Fisher analysis, we find that constraints from persistent homology are tighter for 8 out of the 10 parameters by margins of 13-50%. The complementarity of the two statistics breaks parameter degeneracies, allowing for a further gain in constraining power when combined. We run a series of consistency checks to consolidate our results, and conclude that our findings motivate incorporating persistent homology into inference pipelines for cosmological survey data., Comment: 24+18 pages, 22 figures, 4 tables. Accepted for publication in JCAP. Replaced with the accepted version (minor changes)

Large language models (LLMs) have been shown to exhibit a wide range of capabilities, such as writing robot code from language commands -- enabling non-experts to direct robot behaviors, modify them based on feedback, or compose them to perform new tasks. However, these capabilities (driven by in-context learning) are limited to short-term interactions, where users' feedback remains relevant for only as long as it fits within the context size of the LLM, and can be forgotten over longer interactions. In this work, we investigate fine-tuning the robot code-writing LLMs, to remember their in-context interactions and improve their teachability i.e., how efficiently they adapt to human inputs (measured by average number of corrections before the user considers the task successful). Our key observation is that when human-robot interactions are viewed as a partially observable Markov decision process (in which human language inputs are observations, and robot code outputs are actions), then training an LLM to complete previous interactions is training a transition dynamics model -- that can be combined with classic robotics techniques such as model predictive control (MPC) to discover shorter paths to success. This gives rise to Language Model Predictive Control (LMPC), a framework that fine-tunes PaLM 2 to improve its teachability on 78 tasks across 5 robot embodiments -- improving non-expert teaching success rates of unseen tasks by 26.9% while reducing the average number of human corrections from 2.4 to 1.9. Experiments show that LMPC also produces strong meta-learners, improving the success rate of in-context learning new tasks on unseen robot embodiments and APIs by 31.5%. See videos, code, and demos at: https://robot-teaching.github.io/.

Vision language models (VLMs) have shown impressive capabilities across a variety of tasks, from logical reasoning to visual understanding. This opens the door to richer interaction with the world, for example robotic control. However, VLMs produce only textual outputs, while robotic control and other spatial tasks require outputting continuous coordinates, actions, or trajectories. How can we enable VLMs to handle such settings without fine-tuning on task-specific data? In this paper, we propose a novel visual prompting approach for VLMs that we call Prompting with Iterative Visual Optimization (PIVOT), which casts tasks as iterative visual question answering. In each iteration, the image is annotated with a visual representation of proposals that the VLM can refer to (e.g., candidate robot actions, localizations, or trajectories). The VLM then selects the best ones for the task. These proposals are iteratively refined, allowing the VLM to eventually zero in on the best available answer. We investigate PIVOT on real-world robotic navigation, real-world manipulation from images, instruction following in simulation, and additional spatial inference tasks such as localization. We find, perhaps surprisingly, that our approach enables zero-shot control of robotic systems without any robot training data, navigation in a variety of environments, and other capabilities. Although current performance is far from perfect, our work highlights potentials and limitations of this new regime and shows a promising approach for Internet-Scale VLMs in robotic and spatial reasoning domains. Website: pivot-prompt.github.io and HuggingFace: https://huggingface.co/spaces/pivot-prompt/pivot-prompt-demo.

Compared to Full-Model Fine-Tuning (FMFT), Parameter Efficient Fine-Tuning (PEFT) has demonstrated superior performance and lower computational overhead in several code understanding tasks, such as code summarization and code search. This advantage can be attributed to PEFT's ability to alleviate the catastrophic forgetting issue of Pre-trained Language Models (PLMs) by updating only a small number of parameters. As a result, PEFT effectively harnesses the pre-trained general-purpose knowledge for downstream tasks. However, existing studies primarily involve static code comprehension, aligning with the pre-training paradigm of recent PLMs and facilitating knowledge transfer, but they do not account for dynamic code changes. Thus, it remains unclear whether PEFT outperforms FMFT in task-specific adaptation for code-change-related tasks. To address this question, we examine two prevalent PEFT methods, namely Adapter Tuning (AT) and Low-Rank Adaptation (LoRA), and compare their performance with FMFT on five popular PLMs. Specifically, we evaluate their performance on two widely-studied code-change-related tasks: Just-In-Time Defect Prediction (JIT-DP) and Commit Message Generation (CMG). The results demonstrate that both AT and LoRA achieve state-of-the-art (SOTA) results in JIT-DP and exhibit comparable performances in CMG when compared to FMFT and other SOTA approaches. Furthermore, AT and LoRA exhibit superiority in cross-lingual and low-resource scenarios. We also conduct three probing tasks to explain the efficacy of PEFT techniques on JIT-DP and CMG tasks from both static and dynamic perspectives. The study indicates that PEFT, particularly through the use of AT and LoRA, offers promising advantages in code-change-related tasks, surpassing FMFT in certain aspects., Comment: 12 pages, 5 figures, accepted by SANER 2024 (International Conference on Software Analysis, Evolution, and Reengineering)

We present a tool that leverages generative AI to accelerate the migration of on-premises applications to the cloud. The Cloud Migration LLM accepts input from the user specifying the parameters of their migration, and outputs a migration strategy with an architecture diagram. A user study suggests that the migration LLM can assist inexperienced users in finding the right cloud migration profile, while avoiding complexities of a manual approach., Comment: arXiv admin comment: This version has been removed by arXiv administrators as the submitter did not have the rights to agree to the license at the time of submission

Parallel Reinforcement Learning (RL) frameworks are essential for mapping RL workloads to multiple computational resources, allowing for faster generation of samples, estimation of values, and policy improvement. These computational paradigms require a seamless integration of training, serving, and simulation workloads. Existing frameworks, such as Ray, are not managing this orchestration efficiently, especially in RL tasks that demand intensive input/output and synchronization between actors on a single node. In this study, we have proposed a solution implementing the reactor model, which enforces a set of actors to have a fixed communication pattern. This allows the scheduler to eliminate work needed for synchronization, such as acquiring and releasing locks for each actor or sending and processing coordination-related messages. Our framework, Lingua Franca (LF), a coordination language based on the reactor model, also supports true parallelism in Python and provides a unified interface that allows users to automatically generate dataflow graphs for RL tasks. In comparison to Ray on a single-node multi-core compute platform, LF achieves 1.21x and 11.62x higher simulation throughput in OpenAI Gym and Atari environments, reduces the average training time of synchronized parallel Q-learning by 31.2%, and accelerates multi-agent RL inference by 5.12x., Comment: 10 pages, 11 figures

Code provides a general syntactic structure to build complex programs and perform precise computations when paired with a code interpreter - we hypothesize that language models (LMs) can leverage code-writing to improve Chain of Thought reasoning not only for logic and arithmetic tasks, but also for semantic ones (and in particular, those that are a mix of both). For example, consider prompting an LM to write code that counts the number of times it detects sarcasm in an essay: the LM may struggle to write an implementation for "detect_sarcasm(string)" that can be executed by the interpreter (handling the edge cases would be insurmountable). However, LMs may still produce a valid solution if they not only write code, but also selectively "emulate" the interpreter by generating the expected output of "detect_sarcasm(string)". In this work, we propose Chain of Code (CoC), a simple yet surprisingly effective extension that improves LM code-driven reasoning. The key idea is to encourage LMs to format semantic sub-tasks in a program as flexible pseudocode that the interpreter can explicitly catch undefined behaviors and hand off to simulate with an LM (as an "LMulator"). Experiments demonstrate that Chain of Code outperforms Chain of Thought and other baselines across a variety of benchmarks; on BIG-Bench Hard, Chain of Code achieves 84%, a gain of 12% over Chain of Thought. In a nutshell, CoC broadens the scope of reasoning questions that LMs can answer by "thinking in code"., Comment: ICML 2024 Oral; Project webpage: https://chain-of-code.github.io

{Introduction: } Dementia is a neurological disorder associated with aging that can cause a loss of cognitive functions, impacting daily life. Alzheimer's disease (AD) is the most common cause of dementia, accounting for 50--70\% of cases, while frontotemporal dementia (FTD) affects social skills and personality. Electroencephalography (EEG) provides an effective tool to study the effects of AD on the brain. {Methods: } In this study, we propose to use shallow neural networks applied to two sets of features: spectral-temporal and functional connectivity using four methods. We compare three supervised machine learning techniques to the CNN models to classify EEG signals of AD / FTD and control cases. We also evaluate different measures of functional connectivity from common EEG frequency bands considering multiple thresholds. {Results and Discussion: } Results showed that the shallow CNN-based models achieved the highest accuracy of 94.54\% with AEC in test dataset when considering all connections, outperforming conventional methods and providing potentially an additional early dementia diagnosis tool. \url{https://doi.org/10.3389%2Ffneur.2023.1270405}