623509 research outputs found
Sort by
Classifying binary black holes from Population III stars with the Einstein Telescope: A machine-learning approach
Get PDFThird-generation (3G) gravitational-wave detectors such as the Einstein Telescope (ET) will observe binary black hole (BBH) mergers at redshifts up to . However, an unequivocal determination of the origin of high-redshift sources will remain uncertain because of the low signal-to-noise ratio (S/N) and poor estimate of their luminosity distance. This study proposes a machine-learning approach to infer the origins of high-redshift BBHs. We specifically differentiate those arising from Population III (Pop. III) stars, which probably are the first progenitors of star-born BBH mergers in the Universe, and those originated from Population I-II (Pop. I-II) stars. We considered a wide range of models that encompass the current uncertainties on Pop. III BBH mergers. We then estimated the parameter errors of the detected sources with ET using the Fisher information-matrix formalism, followed by a classification using XGBoost, which is a machine-learning algorithm based on decision trees. For a set of mock observed BBHs, we provide the probability that they belong to the Pop. III class while considering the parameter errors of each source. In our fiducial model, we accurately identify of the detected BBHs that originate from Pop. III stars with a precision . Our study demonstrates that machine-learning enables us to achieve some pivotal aspects of the ET science case by exploring the origin of individual high-redshift GW observations. We set the basis for further studies, which will integrate additional simulated populations and account for further uncertainties in the population modeling.Published in Astronomy & Astrophysics. 15 pages, 9 Figures and 6 tables. Comments are welcom
How Far Can We Go with Practical Function-Level Program Repair?
Get PDFRecently, multiple Automated Program Repair (APR) techniques based on Large Language Models (LLMs) have been proposed to enhance the repair performance. While these techniques mainly focus on the single-line or hunk-level repair, they face significant challenges in real-world application due to the limited repair task scope and costly statement-level fault localization. However, the more practical function-level APR, which broadens the scope of APR task to fix entire buggy functions and requires only cost-efficient function-level fault localization, remains underexplored. In this paper, we conduct the first comprehensive study of LLM-based function-level APR including investigating the effect of the few-shot learning mechanism and the auxiliary repair-relevant information. Specifically, we adopt six widely-studied LLMs and construct a benchmark in both the Defects4J 1.2 and 2.0 datasets. Our study demonstrates that LLMs with zero-shot learning are already powerful function-level APR techniques, while applying the few-shot learning mechanism leads to disparate repair performance. Moreover, we find that directly applying the auxiliary repair-relevant information to LLMs significantly increases function-level repair performance. Inspired by our findings, we propose an LLM-based function-level APR technique, namely SRepair, which adopts a dual-LLM framework to leverage the power of the auxiliary repair-relevant information for advancing the repair performance. The evaluation results demonstrate that SRepair can correctly fix 300 single-function bugs in the Defects4J dataset, largely surpassing all previous APR techniques by at least 85%, without the need for the costly statement-level fault location information. Furthermore, SRepair successfully fixes 32 multi-function bugs in the Defects4J dataset, which is the first time achieved by any APR technique ever to our best knowledge.https://github.com/GhabiX/SRepair
Reinforcement Learning with Adaptive Regularization for Safe Control of Critical Systems
Get PDFReinforcement Learning (RL) is a powerful method for controlling dynamic systems, but its learning mechanism can lead to unpredictable actions that undermine the safety of critical systems. Here, we propose RL with Adaptive Regularization (RL-AR), an algorithm that enables safe RL exploration by combining the RL policy with a policy regularizer that hard-codes the safety constraints. RL-AR performs policy combination via a focus module, which determines the appropriate combination depending on the state--relying more on the safe policy regularizer for less-exploited states while allowing unbiased convergence for well-exploited states. In a series of critical control applications, we demonstrate that RL-AR not only ensures safety during training but also achieves a return competitive with the standards of model-free RL that disregards safety
A relativistic third-order algebraic diagrammatic construction theory for electron detachment, attachment and excitation problems
Get PDFWe present the theory and implementation of a highly efficient relativistic third-order algebraic diagrammatic construction [ADC(3)] method based on a four-component (4c) Dirac-Coulomb (DC) Hamiltonian for the calculation of ionization potentials (IP), electron affinities (EA), and excitation energies (EE). Benchmarking calculations for IP, EA, and EE were performed on both atomic and molecular systems to assess the accuracy of the newly developed four-component relativistic ADC(3) method. The results show good agreement with the available experimental data. The Hermitian nature of the 4c-ADC(3) Hamiltonian, combined with the perturbative truncation of the wave function, offers significant computational advantages over the standard equation-of-motion coupled-cluster approach, particularly for property calculations. The method\u27s suitability for property calculations is further demonstrated by computing oscillator strengths and excited-state dipole moments for heavy elements
Probing Hidden Dimensions via Muon Lifetime Measurements
Get PDFIn the context of Kaluza-Klein theories, the time dilation of charged particles in an external field depends on the charge in a specific way. Experimental tests are proposed to search for extra dimensions using this distinctive feature.V2. Expanded and detailed version of a paper that received an Honorable Mention in the Gravity Research Foundation 2024 Awards. 12 page
Logarithmic Smoothing for Pessimistic Off-Policy Evaluation, Selection and Learning
Get PDFThis work investigates the offline formulation of the contextual bandit problem, where the goal is to leverage past interactions collected under a behavior policy to evaluate, select, and learn new, potentially better-performing, policies. Motivated by critical applications, we move beyond point estimators. Instead, we adopt the principle of pessimism where we construct upper bounds that assess a policy\u27s worst-case performance, enabling us to confidently select and learn improved policies. Precisely, we introduce novel, fully empirical concentration bounds for a broad class of importance weighting risk estimators. These bounds are general enough to cover most existing estimators and pave the way for the development of new ones. In particular, our pursuit of the tightest bound within this class motivates a novel estimator (LS), that logarithmically smooths large importance weights. The bound for LS is provably tighter than its competitors, and naturally results in improved policy selection and learning strategies. Extensive policy evaluation, selection, and learning experiments highlight the versatility and favorable performance of LS.NeuRIPS \u2724 Spotligh
Navigating the Safety Landscape: Measuring Risks in Finetuning Large Language Models
Get PDFSafety alignment is crucial to ensure that large language models (LLMs) behave in ways that align with human preferences and prevent harmful actions during inference. However, recent studies show that the alignment can be easily compromised through finetuning with only a few adversarially designed training examples. We aim to measure the risks in finetuning LLMs through navigating the LLM safety landscape. We discover a new phenomenon observed universally in the model parameter space of popular open-source LLMs, termed as safety basin : random perturbations to model weights maintain the safety level of the original aligned model within its local neighborhood. However, outside this local region, safety is fully compromised, exhibiting a sharp, step-like drop. This safety basin contrasts sharply with the LLM capability landscape, where model performance peaks at the origin and gradually declines as random perturbation increases. Our discovery inspires us to propose the new VISAGE safety metric that measures the safety in LLM finetuning by probing its safety landscape. Visualizing the safety landscape of the aligned model enables us to understand how finetuning compromises safety by dragging the model away from the safety basin. The LLM safety landscape also highlights the system prompt\u27s critical role in protecting a model, and that such protection transfers to its perturbed variants within the safety basin. These observations from our safety landscape research provide new insights for future work on LLM safety community. Our code is publicly available at https://github.com/ShengYun-Peng/llm-landscape.NeurIPS\u272
Task-Agnostic Machine-Learning-Assisted Inference
Get PDFMachine learning (ML) is playing an increasingly important role in scientific research. In conjunction with classical statistical approaches, ML-assisted analytical strategies have shown great promise in accelerating research findings. This has also opened a whole field of methodological research focusing on integrative approaches that leverage both ML and statistics to tackle data science challenges. One type of study that has quickly gained popularity employs ML to predict unobserved outcomes in massive samples, and then uses predicted outcomes in downstream statistical inference. However, existing methods designed to ensure the validity of this type of post-prediction inference are limited to very basic tasks such as linear regression analysis. This is because any extension of these approaches to new, more sophisticated statistical tasks requires task-specific algebraic derivations and software implementations, which ignores the massive library of existing software tools already developed for the same scientific problem given observed data. This severely constrains the scope of application for post-prediction inference. To address this challenge, we introduce a novel statistical framework named PSPS for task-agnostic ML-assisted inference. It provides a post-prediction inference solution that can be easily plugged into almost any established data analysis routines. It delivers valid and efficient inference that is robust to arbitrary choice of ML model, allowing nearly all existing statistical frameworks to be incorporated into the analysis of ML-predicted data. Through extensive experiments, we showcase our method\u27s validity, versatility, and superiority compared to existing approaches. Our software is available at https://github.com/qlu-lab/psps
Visual place recognition for aerial imagery: A survey
Get PDFAerial imagery and its direct application to visual localization is an essential problem for many Robotics and Computer Vision tasks. While Global Navigation Satellite Systems (GNSS) are the standard default solution for solving the aerial localization problem, it is subject to a number of limitations, such as, signal instability or solution unreliability that make this option not so desirable. Consequently, visual geolocalization is emerging as a viable alternative. However, adapting Visual Place Recognition (VPR) task to aerial imagery presents significant challenges, including weather variations and repetitive patterns. Current VPR reviews largely neglect the specific context of aerial data. This paper introduces a methodology tailored for evaluating VPR techniques specifically in the domain of aerial imagery, providing a comprehensive assessment of various methods and their performance. However, we not only compare various VPR methods, but also demonstrate the importance of selecting appropriate zoom and overlap levels when constructing map tiles to achieve maximum efficiency of VPR algorithms in the case of aerial imagery. The code is available on our GitHub repository -- https://github.com/prime-slam/aero-vloc
Sum-Frequency Generation Spectro-Microscopy in the Reststrahlen Band of Wurtzite-type Aluminum Nitride
Get PDFNonlinear-optical microscopy and spectroscopy provide detailed spatial and spectroscopic contrast, specifically sensitive to structural symmetry and order. Ferroics, in particular, have been widely studied using second harmonic generation imaging, which provides detailed information on domain structures but typically lacks spectroscopic detail. In contrast, infrared-visible sum-frequency generation (SFG) spectroscopy reveals details of the atomic structure and bonding via vibrational resonances, but conventionally lacks spatial information. In this work, we combine the benefits of nonlinear optical imaging and SFG spectroscopy by employing SFG spectro-microscopy using an infrared free-electron laser. Specifically, we demonstrate the feasibility of SFG spectro-microscopy for spectroscopy using in-plane anisotropic wurtzite-type aluminum nitride as a model system. We find the experimental spectra to agree well with our theoretical calculations and we show the potential of our microscope to provide spatially resolved spectroscopic information in inhomogeneous systems such as ferroics and their domains in the near future.6 pages, 3 figures, SI upon request to the author