Predicting the collaboration likelihood and measuring cognitive trust to AI systems is more important than ever. To do that, previous research mostly focus solely on the model features (e.g., accuracy, confidence) and ignore the human factor. To address that, we propose several decision-making similarity measures based on divergence metrics (e.g., KL, JSD) calculated over the labels acquired from humans and a wide range of models. We conduct a user study on a textual entailment task, where the users are provided with soft labels from various models and asked to pick the closest option to them. The users are then shown the similarities/differences to their most similar model and are surveyed for their likelihood of collaboration and cognitive trust to the selected system. Finally, we qualitatively and quantitatively analyze the relation between the proposed decision-making similarity measures and the survey results. We find that people tend to collaborate with their most similar models -- measured via JSD -- yet this collaboration does not necessarily imply a similar level of cognitive trust. We release all resources related to the user study (e.g., design, outputs), models, and metrics at our repo.
相關內容
Concerning classical computational models able to express all the Primitive Recursive Functions (PRF), there are interesting results regarding limits on their algorithmic expressiveness or, equivalently, efficiency, namely the ability to express algorithms with minimal computational cost. By introducing the reversible programming model Forest, at our knowledge, we provide a first study of analogous properties, adapted to the context of reversible computational models that can represent all the functions in PRF. Firstly, we show that Forest extends Matos' linear reversible computational model MSRL, the very extension being a guaranteed terminating iteration that can be halted by means of logical predicates. The consequence is that Forest is PRF complete, because MSRL is. Secondly, we show that Forest is strictly algorithmically more expressive than MSRL: it can encode a reversible algorithm for the minimum between two integers in optimal time, while MSRL cannot.
In today's life, more and more people tend to opt for a smart house. In this way, the idea of including technology has become popular worldwide. Despite this concept's many benefits, managing security remains an essential problem due to the shared activities. The Internet of Things system behind a smart house is based on several sensors to measure temperature, humidity, air quality, and movement. Because of being supervised every day through sensors and controlling their house only with a simple click, many people can be afraid of this new approach in terms of their privacy, and this fact can constrain them from following their habits. The security aspects should be constantly analyzed to keep the data's confidentiality and make people feel safe in their own houses. In this context, the current paper puts light on an alternative design of a platform in which the safety of homeowners is the primary purpose, and they maintain complete control over the data generated by smart devices. The current research highlights the role of security and interface design in controlling a smart house. The study underscores the importance of providing an interface that can be used easily by any person to manage data and live activities in a modern residence in an era dominated by continuously developing technology.
As more and more AI agents are used in practice, it is time to think about how to make these agents fully autonomous so that they can learn by themselves in a self-motivated and self-supervised manner rather than being retrained periodically on the initiation of human engineers using expanded training data. As the real-world is an open environment with unknowns or novelties, detecting novelties or unknowns, characterizing them, accommodating or adapting to them, gathering ground-truth training data, and incrementally learning the unknowns/novelties are critical to making the agent more and more knowledgeable and powerful over time. The key challenge is how to automate the process so that it is carried out on the agent's own initiative and through its own interactions with humans and the environment. Since an AI agent usually has a performance task, characterizing each novelty becomes critical and necessary so that the agent can formulate an appropriate response to adapt its behavior to accommodate the novelty and to learn from it to improve the agent's adaptation capability and task performance. The process goes continually without termination. This paper proposes a theoretic framework for this learning paradigm to promote the research of building Self-initiated Open world Learning (SOL) agents. An example SOL agent is also described.
This paper studies the measurement of advertising effects on online platforms when parallel experimentation occurs, that is, when multiple advertisers experiment concurrently. It provides a framework that makes precise how parallel experimentation affects the experiment's value: while ignoring parallel experimentation yields an estimate of the average effect of advertising in-place, which has limited value in decision-making in an environment with variable advertising competition, accounting for parallel experimentation captures the actual uncertainty advertisers face due to competitive actions. It then implements an experimental design that enables the estimation of these effects on JD.com, a large e-commerce platform that is also a publisher of digital ads. Using traditional and kernel-based estimators, it shows that not accounting for competitive actions can result in the advertiser inaccurately estimating the advertising lift by a factor of two or higher, which can be consequential for decision-making.
This paper generalizes the notion of sufficiency for estimation problems beyond maximum likelihood. In particular, we consider estimation problems based on Jones et al. and Basu et al. likelihood functions that are popular among distance-based robust inference methods. We first characterize the probability distributions that always have a fixed number of sufficient statistics (independent of sample size) with respect to these likelihood functions. These distributions are power-law extensions of the usual exponential family and contain Student distributions as a special case. We then extend the notion of minimal sufficient statistics and compute it for these power-law families. Finally, we establish a Rao-Blackwell-type theorem for finding the best estimators for a power-law family. This helps us establish Cram\'er-Rao-type lower bounds for power-law families.
As service robots become more capable of autonomous behaviors, it becomes increasingly important to consider how people communicate with a robot what task it should perform and how to do the task. Accordingly, there has been a rise in attention to end-user development (EUD) interfaces, which enable non-roboticist end users to specify tasks for autonomous robots to perform. However, state-of-the-art EUD interfaces are often constrained through simplified domains or restrictive end-user interaction. Motivated by prior qualitative design work that explores how to integrate a care robot in an assisted living community, we discuss the challenges of EUD in this complex domain. One set of challenges stems from different user-facing representations, e.g., certain tasks may lend themselves better to rule-based trigger-action representations, whereas other tasks may be easier to specify via sequences of actions. The other stems from considering the needs of multiple stakeholders, e.g., caregivers and residents of the facility may all create tasks for the robot, but the robot may not be able to share information about all tasks with all residents due to privacy concerns. We present scenarios that illustrate these challenges and also discuss possible solutions.
As research and deployment of AI grows, the computational burden to support and sustain its progress inevitably does too. To train or fine-tune state-of-the-art models in NLP, computer vision, etc., some form of AI hardware acceleration is virtually a requirement. Recent large language models require considerable resources to train and deploy, resulting in significant energy usage, potential carbon emissions, and massive demand for GPUs and other hardware accelerators. However, this surge carries large implications for energy sustainability at the HPC/datacenter level. In this paper, we study the aggregate effect of power-capping GPUs on GPU temperature and power draw at a research supercomputing center. With the right amount of power-capping, we show significant decreases in both temperature and power draw, reducing power consumption and potentially improving hardware life-span with minimal impact on job performance. While power-capping reduces power draw by design, the aggregate system-wide effect on overall energy consumption is less clear; for instance, if users notice job performance degradation from GPU power-caps, they may request additional GPU-jobs to compensate, negating any energy savings or even worsening energy consumption. To our knowledge, our work is the first to conduct and make available a detailed analysis of the effects of GPU power-capping at the supercomputing scale. We hope our work will inspire HPCs/datacenters to further explore, evaluate, and communicate the impact of power-capping AI hardware accelerators for more sustainable AI.
This article presents the affordances that Generative Artificial Intelligence can have in disinformation context, one of the major threats to our digitalized society. We present a research framework to generate customized agent-based social networks for disinformation simulations that would enable understanding and evaluation of the phenomena whilst discussing open challenges.
It has been shown that deep neural networks are prone to overfitting on biased training data. Towards addressing this issue, meta-learning employs a meta model for correcting the training bias. Despite the promising performances, super slow training is currently the bottleneck in the meta learning approaches. In this paper, we introduce a novel Faster Meta Update Strategy (FaMUS) to replace the most expensive step in the meta gradient computation with a faster layer-wise approximation. We empirically find that FaMUS yields not only a reasonably accurate but also a low-variance approximation of the meta gradient. We conduct extensive experiments to verify the proposed method on two tasks. We show our method is able to save two-thirds of the training time while still maintaining the comparable or achieving even better generalization performance. In particular, our method achieves the state-of-the-art performance on both synthetic and realistic noisy labels, and obtains promising performance on long-tailed recognition on standard benchmarks.
Deep neural networks (DNNs) are successful in many computer vision tasks. However, the most accurate DNNs require millions of parameters and operations, making them energy, computation and memory intensive. This impedes the deployment of large DNNs in low-power devices with limited compute resources. Recent research improves DNN models by reducing the memory requirement, energy consumption, and number of operations without significantly decreasing the accuracy. This paper surveys the progress of low-power deep learning and computer vision, specifically in regards to inference, and discusses the methods for compacting and accelerating DNN models. The techniques can be divided into four major categories: (1) parameter quantization and pruning, (2) compressed convolutional filters and matrix factorization, (3) network architecture search, and (4) knowledge distillation. We analyze the accuracy, advantages, disadvantages, and potential solutions to the problems with the techniques in each category. We also discuss new evaluation metrics as a guideline for future research.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191