fairness

"Moral alignment is defined as the perceived congruence between the values embedded in an AI system's decision logic and the moral intuitions of stakeholders."

"Also, an interesting finding that training for automated labels may amplify bias in your model by 40%."

"This paper introduces an online-within-online fair multi-task learning (OWO-FMTL) framework that ensures long-term equity across users."

"Taikyo (https://taikyo.xyz) is a service that performs an automatic diagnosis by simply uploading the itemized statement of moving-out expenses, in accordance with the guidelines of the Ministry of Land, Infrastructure, Transport and Tourism."

"一方、本論文で提案された Average Bias-Boundedness (A-BB) は、バイアスを数理的に定義し、その上限を理論的に保証しながら評価を行う枠組みです。"

"The results indicate measurable improvements in key interaction qualities, with average user ratings of 4.7 out of 5 for politeness and 4.9 out of 5 for fairness."

"Our experiments reveal that label-clustered CP variants consistently deliver superior substantive fairness."

"Microsoft is investigating the exploitation of AI's memory functions to favor specific companies."

"Experimental results demonstrate the effectiveness of these methods in (i) guiding ML models to focus on relevant image features, particularly when CAIPI is used, and (ii) reducing model bias (i.e., balancing the misclassification rates between male and female predictions)."

"We propose a novel double fairness learning (DFL) framework that explicitly manages the trade-off among three objectives: action fairness, outcome fairness, and value maximization."

"Through a user study with participants across the United States, India, Kenya, and Nigeria, we investigate whether LLM-simulated users serve as reliable proxies for real human users in evaluating agents on { au}-Bench retail tasks."

"These findings indicate that AI systems is still imperfectly trained on noisy, sometimes non-causal signals that do not reliably reflect true patient acuity."

"I dunno what to flair this - it’s not like I prompted it to be sexist or anything, it just… is"

"AI bias is a core theme in AI ethics because it can lead to real-world harms such as discrimination, unfairness, and loss of opportunity."

"Assuming the article argues that AI 'slop' originates from human input: "The garbage in, garbage out principle applies directly to AI training.""

"The context mentions the source of the article is ArXiv."

"The paper focuses on removing bias without erasing demographics."

"The context mentions the article is sourced from ArXiv, indicating it is a pre-print research paper."

"The study focuses on empirically characterizing tone bias in LLM-driven UX systems."

"The research focuses on 93 stigmatized groups."

"The research focuses on the performance trade-offs associated with mitigating bias."

"The article focuses on fair regression for multiple groups in the context of Chronic Kidney Disease."

"The paper is available on ArXiv, suggesting peer review is not yet complete but the research is publicly accessible."

"The paper focuses on bias and discrimination in memory-enhanced AI agents."

"The article likely explores emergent bias and fairness within the context of multi-agent decision systems."

"The study focuses on cross-language bias."

"The paper focuses on vision-language models for medical image disease classification."

"The research focuses on the bias of the Gini estimator in Poisson and geometric cases, also characterizing the gamma family and unbiasedness under gamma distributions."

"The study utilizes interviews to gather insights."

"The study explores the alignment of language models with specific cultural values and the effects of cultural prompting."