arXiv subjects

Hiromi Arai

Publications and source records attributed to Hiromi Arai.

9 recordsLinked to original sources

arXiv2026

Probability Bounding: Post-Hoc Calibration via Box-Constrained Softmax

Many studies have observed that modern neural networks achieve high accuracy while producing poorly calibrated probabilities, making calibration a critical practical issue. In this work, we propose probability bounding (PB), a novel post-hoc calibration method that mitigates both underconfidence and overconfidence by learning lower and upper bounds on the output probabilities. To implement PB, we introduce the box-constrained softmax (BCSoftmax) function, a generalization of Softmax that explicitly enforces lower and upper bounds on the output probabilities. While BCSoftmax is formulated as the solution to a box-constrained optimization problem, we develop an exact and efficient algorithm for computing BCSoftmax. We further provide theoretical guarantees for PB and introduce two variants of PB. We demonstrate the effectiveness of our methods experimentally on four real-world datasets, consistently reducing calibration errors. Our Python implementation is available at https://github.com/neonnnnn/torchbcsoftmax.

stat.ML↗

arXiv2025

JBBQ: Japanese Bias Benchmark for Analyzing Social Biases in Large Language Models

With the development of large language models (LLMs), social biases in these LLMs have become a pressing issue. Although there are various benchmarks for social biases across languages, the extent to which Japanese LLMs exhibit social biases has not been fully investigated. In this study, we construct the Japanese Bias Benchmark dataset for Question Answering (JBBQ) based on the English bias benchmark BBQ, with analysis of social biases in Japanese LLMs. The results show that while current open Japanese LLMs with more parameters show improved accuracies on JBBQ, their bias scores increase. In addition, prompts with a warning about social biases and chain-of-thought prompting reduce the effect of biases in model outputs, but there is room for improvement in extracting the correct evidence from contexts in Japanese. Our dataset is available at https://github.com/ynklab/JBBQ_data.

cs.CL↗

arXiv2024

Fair Machine Guidance to Enhance Fair Decision Making in Biased People

Teaching unbiased decision-making is crucial for addressing biased decision-making in daily life. Although both raising awareness of personal biases and providing guidance on unbiased decision-making are essential, the latter topics remains under-researched. In this study, we developed and evaluated an AI system aimed at educating individuals on making unbiased decisions using fairness-aware machine learning. In a between-subjects experimental design, 99 participants who were prone to bias performed personal assessment tasks. They were divided into two groups: a) those who received AI guidance for fair decision-making before the task and b) those who received no such guidance but were informed of their biases. The results suggest that although several participants doubted the fairness of the AI system, fair machine guidance prompted them to reassess their views regarding fairness, reflect on their biases, and modify their decision-making criteria. Our findings provide insights into the design of AI systems for guiding fair decision-making in humans.

cs.HC↗

arXiv2023

Will Large-scale Generative Models Corrupt Future Datasets?

Recently proposed large-scale text-to-image generative models such as DALL$\cdot$E 2, Midjourney, and StableDiffusion can generate high-quality and realistic images from users' prompts. Not limited to the research community, ordinary Internet users enjoy these generative models, and consequently, a tremendous amount of generated images have been shared on the Internet. Meanwhile, today's success of deep learning in the computer vision field owes a lot to images collected from the Internet. These trends lead us to a research question: "\textbf{will such generated images impact the quality of future datasets and the performance of computer vision models positively or negatively?}" This paper empirically answers this question by simulating contamination. Namely, we generate ImageNet-scale and COCO-scale datasets using a state-of-the-art generative model and evaluate models trained with "contaminated" datasets on various tasks, including image classification and image generation. Throughout experiments, we conclude that generated images negatively affect downstream performance, while the significance depends on tasks and the amount of generated images. The generated datasets and the codes for experiments will be publicly released for future research. Generated datasets and source codes are available from \url{https://github.com/moskomule/dataset-contamination}.

cs.CV↗

arXiv2022

Trustworthy Human Computation: A Survey

Human computation is an approach to solving problems that prove difficult using AI only, and involves the cooperation of many humans. Because human computation requires close engagement with both "human populations as users" and "human populations as driving forces," establishing mutual trust between AI and humans is an important issue to further the development of human computation. This survey lays the groundwork for the realization of trustworthy human computation. First, the trustworthiness of human computation as computing systems, that is, trust offered by humans to AI, is examined using the RAS (Reliability, Availability, and Serviceability) analogy, which define measures of trustworthiness in conventional computer systems. Next, the social trustworthiness provided by human computation systems to users or participants is discussed from the perspective of AI ethics, including fairness, privacy, and transparency. Then, we consider human--AI collaboration based on two-way trust, in which humans and AI build mutual trust and accomplish difficult tasks through reciprocal collaboration. Finally, future challenges and research directions for realizing trustworthy human computation are discussed.

cs.AI↗

arXiv2022

Designing a Location Trace Anonymization Contest

For a better understanding of anonymization methods for location traces, we have designed and held a location trace anonymization contest that deals with a long trace (400 events per user) and fine-grained locations (1024 regions). In our contest, each team anonymizes her original traces, and then the other teams perform privacy attacks against the anonymized traces. In other words, both defense and attack compete together, which is close to what happens in real life. Prior to our contest, we show that re-identification alone is insufficient as a privacy risk and that trace inference should be added as an additional risk. Specifically, we show an example of anonymization that is perfectly secure against re-identification and is not secure against trace inference. Based on this, our contest evaluates both the re-identification risk and trace inference risk and analyzes their relationship. Through our contest, we show several findings in a situation where both defense and attack compete together. In particular, we show that an anonymization method secure against trace inference is also secure against re-identification under the presence of appropriate pseudonymization. We also report defense and attack algorithms that won first place, and analyze the utility of anonymized traces submitted by teams in various applications such as POI recommendation and geo-data analysis.

cs.CR↗

arXiv2021

Characterizing the risk of fairwashing

Fairwashing refers to the risk that an unfair black-box model can be explained by a fairer model through post-hoc explanation manipulation. In this paper, we investigate the capability of fairwashing attacks by analyzing their fidelity-unfairness trade-offs. In particular, we show that fairwashed explanation models can generalize beyond the suing group (i.e., data points that are being explained), meaning that a fairwashed explainer can be used to rationalize subsequent unfair decisions of a black-box model. We also demonstrate that fairwashing attacks can transfer across black-box models, meaning that other black-box models can perform fairwashing without explicitly using their predictions. This generalization and transferability of fairwashing attacks imply that their detection will be difficult in practice. Finally, we propose an approach to quantify the risk of fairwashing, which is based on the computation of the range of the unfairness of high-fidelity explainers.

cs.LG↗

arXiv2020

Ethical Issues Regarding the Use of AI Profiling Services for Recruiting: The Japanese Rikunabi Data Scandal

The ethical, legal, and social challenges involved in the use of profiling services for recruitment are the focus of many previous studies; however, the processes vary depending on the social system and cultural practices. In August 2019, a scandal occurred in Japan in which a recruitment management company was found to have breached users' and students' trust by selling their data to clients. By sharing the Japanese recruitment context and associated laws, this article contributes to our understanding of the ethical issues involved in artificial intelligence profiling and in handling sensitive personal information.

cs.CY↗

arXiv2019

Fairwashing: the risk of rationalization

Black-box explanation is the problem of explaining how a machine learning model -- whose internal logic is hidden to the auditor and generally complex -- produces its outcomes. Current approaches for solving this problem include model explanation, outcome explanation as well as model inspection. While these techniques can be beneficial by providing interpretability, they can be used in a negative manner to perform fairwashing, which we define as promoting the false perception that a machine learning model respects some ethical values. In particular, we demonstrate that it is possible to systematically rationalize decisions taken by an unfair black-box model using the model explanation as well as the outcome explanation approaches with a given fairness metric. Our solution, LaundryML, is based on a regularized rule list enumeration algorithm whose objective is to search for fair rule lists approximating an unfair black-box model. We empirically evaluate our rationalization technique on black-box models trained on real-world datasets and show that one can obtain rule lists with high fidelity to the black-box model while being considerably less unfair at the same time.

cs.LG↗