623509 research outputs found
Sort by
When Quantum Meets Classical: Characterizing Hybrid Quantum-Classical Issues Discussed in Developer Forums
No full textRecent advances in quantum computing have sparked excitement that this new computing paradigm could solve previously intractable problems. However, due to the faulty nature of current quantum hardware and quantum-intrinsic noise, the full potential of quantum computing is still years away. Hybrid quantum-classical computing has emerged as a possible compromise that achieves the best of both worlds. In this paper, we look at hybrid quantum-classical computing from a software engineering perspective and present the first empirical study focused on characterizing and evaluating recurrent issues faced by developers of hybrid quantum-classical applications. The study comprised a thorough analysis of 531 real-world issues faced by developers -- including software faults, hardware failures, quantum library errors, and developer mistakes -- documented in discussion threads from forums dedicated to quantum computing. By qualitatively analyzing such forum threads, we derive a comprehensive taxonomy of recurring issues in hybrid quantum-classical applications that can be used by both application and platform developers to improve the reliability of hybrid applications. The study considered how these recurring issues manifest and their causes, determining that hybrid applications are crash-dominant (74% of studied issues) and that errors were predominantly introduced by application developers (70% of issues). We conclude by identifying recurring obstacles for developers of hybrid applications and actionable recommendations to overcome them
Hierarchical Prompt Decision Transformer: Improving Few-Shot Policy Generalization with Global and Adaptive Guidance
No full textDecision transformers recast reinforcement learning as a conditional sequence generation problem, offering a simple but effective alternative to traditional value or policy-based methods. A recent key development in this area is the integration of prompting in decision transformers to facilitate few-shot policy generalization. However, current methods mainly use static prompt segments to guide rollouts, limiting their ability to provide context-specific guidance. Addressing this, we introduce a hierarchical prompting approach enabled by retrieval augmentation. Our method learns two layers of soft tokens as guiding prompts: (1) global tokens encapsulating task-level information about trajectories, and (2) adaptive tokens that deliver focused, timestep-specific instructions. The adaptive tokens are dynamically retrieved from a curated set of demonstration segments, ensuring context-aware guidance. Experiments across seven benchmark tasks in the MuJoCo and MetaWorld environments demonstrate the proposed approach consistently outperforms all baseline methods, suggesting that hierarchical prompting for decision transformers is an effective strategy to enable few-shot policy generalization
Orbit determination of visual binary systems observed with CST telescope in 2010-2013
Get PDFWe present new orbital solutions for 15 binaries, which were astrometrically measured during 2010-2013. We observe the binary systems using the FastCam ``lucky-imaging\u27\u27 camera, installed at the 1.5-m Carlos Sánchez Telescope (CST) at the Observatorio del Teide, Tenerife (Spain). We present first orbital solutions for four binaries and revise orbits for other 11 systems. We apply two orbital calculation techniques, the ``three-dimensional grid search method\u27\u27 and the Docobo\u27s analytical method. In addition, we use our tool ``Binary Deblending\u27\u27, which is based on deblending the entire observed multiband photometry into fundamental and photometric parameters for each stellar component based on PARSEC isochrones. This method allows us to obtain the total mass for all systems. Our findings include the identification of a binary system consisting of two M-type dwarfs (WOR 19), a binary of evolved components (twin F6IV-V stars) in BU 1292, accompanied by a newly discovered wide (10.5 arcsec) and faint companion with G=17.05 mag. Additionally, we explore the X-ray emission system STF 147 and a very young quadruple system, WDS 04573+5345. This comprehensive analysis significantly contributes to our understanding on the formation and evolution of these stellar systems.To be published in PASA journal. 18 pages, 4 figures containing several plots and 6 table
Assessing Personalized AI Mentoring with Large Language Models in the Computing Field
No full textThis paper provides an in-depth evaluation of three state-of-the-art Large Language Models (LLMs) for personalized career mentoring in the computing field, using three distinct student profiles that consider gender, race, and professional levels. We evaluated the performance of GPT-4, LLaMA 3, and Palm 2 using a zero-shot learning approach without human intervention. A quantitative evaluation was conducted through a custom natural language processing analytics pipeline to highlight the uniqueness of the responses and to identify words reflecting each student\u27s profile, including race, gender, or professional level. The analysis of frequently used words in the responses indicates that GPT-4 offers more personalized mentoring compared to the other two LLMs. Additionally, a qualitative evaluation was performed to see if human experts reached similar conclusions. The analysis of survey responses shows that GPT-4 outperformed the other two LLMs in delivering more accurate and useful mentoring while addressing specific challenges with encouragement languages. Our work establishes a foundation for developing personalized mentoring tools based on LLMs, incorporating human mentors in the process to deliver a more impactful and tailored mentoring experience
Radiology Report Generation via Multi-objective Preference Optimization
Get PDFAutomatic Radiology Report Generation (RRG) is an important topic for alleviating the substantial workload of radiologists. Existing RRG approaches rely on supervised regression based on different architectures or additional knowledge injection,while the generated report may not align optimally with radiologists\u27 preferences. Especially, since the preferences of radiologists are inherently heterogeneous and multidimensional, e.g., some may prioritize report fluency, while others emphasize clinical accuracy. To address this problem,we propose a new RRG method via Multi-objective Preference Optimization (MPO) to align the pre-trained RRG model with multiple human preferences, which can be formulated by multi-dimensional reward functions and optimized by multi-objective reinforcement learning (RL). Specifically, we use a preference vector to represent the weight of preferences and use it as a condition for the RRG model. Then, a linearly weighed reward is obtained via a dot product between the preference vector and multi-dimensional reward. Next,the RRG model is optimized to align with the preference vector by optimizing such a reward via RL. In the training stage,we randomly sample diverse preference vectors from the preference space and align the model by optimizing the weighted multi-objective rewards, which leads to an optimal policy on the entire preference space. When inference,our model can generate reports aligned with specific preferences without further fine-tuning. Extensive experiments on two public datasets show the proposed method can generate reports that cater to different preferences in a single model and achieve state-of-the-art performance.Accepted by AAAI 202
Hidden Biases of End-to-End Driving Datasets
No full textEnd-to-end driving systems have made rapid progress, but have so far not been applied to the challenging new CARLA Leaderboard 2.0. Further, while there is a large body of literature on end-to-end architectures and training strategies, the impact of the training dataset is often overlooked. In this work, we make a first attempt at end-to-end driving for Leaderboard 2.0. Instead of investigating architectures, we systematically analyze the training dataset, leading to new insights: (1) Expert style significantly affects downstream policy performance. (2) In complex data sets, the frames should not be weighted on the basis of simplistic criteria such as class frequencies. (3) Instead, estimating whether a frame changes the target labels compared to previous frames can reduce the size of the dataset without removing important information. By incorporating these findings, our model ranks first and second respectively on the map and sensors tracks of the 2024 CARLA Challenge, and sets a new state-of-the-art on the Bench2Drive test routes. Finally, we uncover a design flaw in the current evaluation metrics and propose a modification for future challenges. Our dataset, code, and pre-trained models are publicly available at https://github.com/autonomousvision/carla_garage.Technical report for the CVPR 2024 Workshop on Foundation Models for Autonomous Systems. Runner-up of the track \u27CARLA Autonomous Driving Challenge\u27 in the 2024 Autonomous Grand Challenge (https://opendrivelab.com/challenge2024/
NERsocial: Efficient Named Entity Recognition Dataset Construction for Human-Robot Interaction Utilizing RapidNER
No full textAdapting named entity recognition (NER) methods to new domains poses significant challenges. We introduce RapidNER, a framework designed for the rapid deployment of NER systems through efficient dataset construction. RapidNER operates through three key steps: (1) extracting domain-specific sub-graphs and triples from a general knowledge graph, (2) collecting and leveraging texts from various sources to build the NERsocial dataset, which focuses on entities typical in human-robot interaction, and (3) implementing an annotation scheme using Elasticsearch (ES) to enhance efficiency. NERsocial, validated by human annotators, includes six entity types, 153K tokens, and 99.4K sentences, demonstrating RapidNER\u27s capability to expedite dataset creation
On study of transition fronts of Fisher-KPP type reaction-diffusion PDEs by non-linear transformations into exactly solvable class
No full textSpatio-temporal dynamics of the evolution of population involving growth and diffusion processes can be modeled by class of partial diffusion equations (PDEs) known as reaction-diffusion systems. In this work, we developed a nonlinear transformations method that converts the original nonlinear Fisher-KPP class of PDEs into an exactly solvable class. We then demonstrated that the proposed nonlinear transformation method intrinsically preserves the relaxation behavior of the solutions to asymptotic values of the non-linear dynamical system. We also show that these particular transforms are very amenable to yield an exact closed form solution in terms of the heat kernel and analytical approximations through the two variable Hermite polynomials. With this proposed method, we calculated the front velocity and shape of the propagating wave and showed how the non-linear transformation affects these parameters for both short and long epochs. As applications, we focus on solving pertinent cases of the Fisher-KPP type of PDEs relating to the evolutionary dynamics by assigning fitness to the mutant gene according to zygosity conditions. We calculated the relaxation of velocity with the parameters of the initial conditions in the following cases, namely, the Fisher, the heterozygote inferior fitness, the heterozygote superior fitness, and finally a general nonlinearity case. We also verified previous conjectures through the exact solutions computed using the proposed method.39 pages, 10 figure
Milky Way dark matter distribution or MOND test from vertical stellar kinematics with Gaia DR3
No full textVertical stellar kinematics+density can be used to trace the dark matter distribution [or the equivalent phantom mass in a Modified Newtonian Dynamics (MOND) scenario] through Jeans equations. In this paper, we want to improve this type of analysis by making use of the recent data of the 6D information from the Gaia-DR3 survey in the anticenter and the Galactic poles to obtain the dynamical mass distribution near plane regions, including extended kinematics over a wide region of 8 kpc22 kpc, kpc.
Our conclusions are as follows: (i) the model of the spherical dark matter halos and the MOND model are compatible with the data; (ii) the model of the disky dark matter (with density proportional to the gas density) is excluded; (iii) the total lack of dark matter (there is only visible matter) within Newtonian gravity is compatible with the data; for instance, at solar Galactocentric radius, we obtained M pc for kpc, compatible with the expected value for visible matter alone of 44 M pc, thus allowing zero dark matter. Similarly, for , kpc: , , , M pc, respectively, for kpc, compatible with visible matter alone. Larger error bars in comparison with previous works are not due to worse data or a more awkward technique but to a stricter modeling of the stellar distribution.21 pages, 15 figures, accepted to be published in Ap
DQA: An Efficient Method for Deep Quantization of Deep Neural Network Activations
Get PDFQuantization of Deep Neural Network (DNN) activations is a commonly used technique to reduce compute and memory demands during DNN inference, which can be particularly beneficial on resource-constrained devices. To achieve high accuracy, existing methods for quantizing activations rely on complex mathematical computations or perform extensive searches for the best hyper-parameters. However, these expensive operations are impractical on devices with limited computation capabilities, memory capacities, and energy budgets. Furthermore, many existing methods do not focus on sub-6-bit (or deep) quantization.
To fill these gaps, in this paper we propose DQA (Deep Quantization of DNN Activations), a new method that focuses on sub-6-bit quantization of activations and leverages simple shifting-based operations and Huffman coding to be efficient and achieve high accuracy. We evaluate DQA with 3, 4, and 5-bit quantization levels and three different DNN models for two different tasks, image classification and image segmentation, on two different datasets. DQA shows significantly better accuracy (up to 29.28%) compared to the direct quantization method and the state-of-the-art NoisyQuant for sub-6-bit quantization.Accepted to Second Workshop on Machine Learning with New Compute Paradigms at NeurIPS 2024 (MLNCP 2024