Vertical Federated Learning (VFL) has emerged as a critical approach in machine learning to address privacy concerns associated with centralized data storage and processing. VFL facilitates collaboration among multiple entities with distinct feature sets on the same user population, enabling the joint training of predictive models without direct data sharing. A key aspect of VFL is the fair and accurate evaluation of each entity's contribution to the learning process. This is crucial for maintaining trust among participating entities, ensuring equitable resource sharing, and fostering a sustainable collaboration framework. This paper provides a thorough review of contribution evaluation in VFL. We categorize the vast array of contribution evaluation techniques along the VFL lifecycle, granularity of evaluation, privacy considerations, and core computational methods. We also explore various tasks in VFL that involving contribution evaluation and analyze their required evaluation properties and relation to the VFL lifecycle phases. Finally, we present a vision for the future challenges of contribution evaluation in VFL. By providing a structured analysis of the current landscape and potential advancements, this paper aims to guide researchers and practitioners in the design and implementation of more effective, efficient, and privacy-centric VFL solutions. Relevant literature and open-source resources have been compiled and are being continuously updated at the GitHub repository: \url{https://github.com/cuiyuebing/VFL_CE}.

This paper introduces "Semantic Scaling," a novel method for ideal point estimation from text. I leverage large language models to classify documents based on their expressed stances and extract survey-like data. I then use item response theory to scale subjects from these data. Semantic Scaling significantly improves on existing text-based scaling methods, and allows researchers to explicitly define the ideological dimensions they measure. This represents the first scaling approach that allows such flexibility outside of survey instruments and opens new avenues of inquiry for populations difficult to survey. Additionally, it works with documents of varying length, and produces valid estimates of both mass and elite ideology. I demonstrate that the method can differentiate between policy preferences and in-group/out-group affect. Among the public, Semantic Scaling out-preforms Tweetscores according to human judgement; in Congress, it recaptures the first dimension DW-NOMINATE while allowing for greater flexibility in resolving construct validity challenges.

We conducted an in-person ethnography in India and the US to investigate how young people (18-24) trusted online content, with a focus on generative AI (GenAI). We had four key findings about how young people use GenAI and determine what to trust online. First, when online, we found participants fluidly shifted between mindsets and emotional states, which we term "information modes." Second, these information modes shaped how and why participants trust GenAI and how they applied literacy skills. In the modes where they spent most of their time, they eschewed literacy skills. Third, with the advent of GenAI, participants imported existing trust heuristics from familiar online contexts into their interactions with GenAI. Fourth, although study participants had reservations about GenAI, they saw it as a requisite tool to adopt to keep up with the times. Participants valued efficiency above all else, and used GenAI to further their goals quickly at the expense of accuracy. Our findings suggest that young people spend the majority of their time online not concerned with truth because they are seeking only to pass the time. As a result, literacy interventions should be designed to intervene at the right time, to match users' distinct information modes, and to work with their existing fact-checking practices.

As the capabilities of Large Language Models (LLMs) in healthcare and medicine continue to advance, there is a growing need for competitive open-source models that can safeguard public interest. With the increasing availability of highly competitive open base models, the impact of continued pre-training is increasingly uncertain. In this work, we explore the role of instruct tuning, model merging, alignment, red teaming and advanced inference schemes, as means to improve current open models. To that end, we introduce the Aloe family, a set of open medical LLMs highly competitive within its scale range. Aloe models are trained on the current best base models (Mistral, LLaMA 3), using a new custom dataset which combines public data sources improved with synthetic Chain of Thought (CoT). Aloe models undergo an alignment phase, becoming one of the first few policy-aligned open healthcare LLM using Direct Preference Optimization, setting a new standard for ethical performance in healthcare LLMs. Model evaluation expands to include various bias and toxicity datasets, a dedicated red teaming effort, and a much-needed risk assessment for healthcare LLMs. Finally, to explore the limits of current LLMs in inference, we study several advanced prompt engineering strategies to boost performance across benchmarks, yielding state-of-the-art results for open healthcare 7B LLMs, unprecedented at this scale.

Apple released its earnings report on Thursday, revealing a drop in overall revenue fueled by slackening iPhone sales. Earnings exceeded market expectations, however, and Apple's shares rose in after-hours trading. Tim Cook, Apple's chief executive, said in a statement released before the call that "Apple is reporting revenue of 90.8bn for the March quarter, including an all-time revenue record in services". The iPhone maker reported revenue of 90.8bn, down 4% year-over-year, but surpassing anticipated earnings of 90.1bn. It declared 0.25 in cash dividend for each share, an increase of 4%.

Emails from the Department of Justice's antitrust case against Google revealed how Microsoft executives were alarmed by and even envious of Google's AI lead. In an email thread, CTO Kevin Scott wrote he was "very, very worried" about Google's rapidly growing AI capabilities. He said he initially dismissed the company's "game-playing stunts," likely referring to Google's AlphaGo models. The emails reference Gmail's autocomplete features, which execs called "scary good." Microsoft struggled to copy Google's BERT-large, an AI model that deciphers the meaning and context of words in a sentence.

To reduce the communication overhead caused by parallel training of multiple clients, various federated learning (FL) techniques use random client sampling. Nonetheless, ensuring the efficacy of random sampling and determining the optimal number of clients to sample in federated recommender systems (FRSs) remains challenging due to the isolated nature of each user as a separate client. This challenge is exacerbated in models where public and private features can be separated, and FL allows communication of only public features (item gradients). In this study, we establish sample complexity bounds that dictate the ideal number of clients required for improved communication efficiency and retained accuracy in such models. In line with our theoretical findings, we empirically demonstrate that RS-FairFRS reduces communication cost (~47%). Second, we demonstrate the presence of class imbalance among clients that raises a substantial equity concern for FRSs. Unlike centralized machine learning, clients in FRS can not share raw data, including sensitive attributes. For this, we introduce RS-FairFRS, first fairness under unawareness FRS built upon random sampling based FRS. While random sampling improves communication efficiency, we propose a novel two-phase dual-fair update technique to achieve fairness without revealing protected attributes of active clients participating in training. Our results on real-world datasets and different sensitive features illustrate a significant reduction in demographic bias (~approx40\%), offering a promising path to achieving fairness and communication efficiency in FRSs without compromising the overall accuracy of FRS.

As AI systems become increasingly prevalent and impactful, the need for effective AI governance and accountability measures is paramount. This paper examines the AI governance landscape, focusing on Anthropic's Claude, a foundational AI model. We analyze Claude through the lens of the NIST AI Risk Management Framework and the EU AI Act, identifying potential threats and proposing mitigation strategies. The paper highlights the importance of transparency, rigorous benchmarking, and comprehensive data handling processes in ensuring the responsible development and deployment of AI systems. We conclude by discussing the social impact of AI governance and the ethical considerations surrounding AI accountability.

This paper presents a set of intersectional feminist principles for conducting equitable, ethical, and sustainable AI research. In Data Feminism (2020), we offered seven principles for examining and challenging unequal power in data science. Here, we present a rationale for why feminism remains deeply relevant for AI research, rearticulate the original principles of data feminism with respect to AI, and introduce two potential new principles related to environmental impact and consent. Together, these principles help to 1) account for the unequal, undemocratic, extractive, and exclusionary forces at work in AI research, development, and deployment; 2) identify and mitigate predictable harms in advance of unsafe, discriminatory, or otherwise oppressive systems being released into the world; and 3) inspire creative, joyful, and collective ways to work towards a more equitable, sustainable world in which all of us can thrive.

Alignment is a standard procedure to fine-tune pre-trained large language models (LLMs) to follow natural language instructions and serve as helpful AI assistants. We have observed, however, that the conventional alignment process fails to enhance the factual accuracy of LLMs, and often leads to the generation of more false facts (i.e. hallucination). In this paper, we study how to make the LLM alignment process more factual, by first identifying factors that lead to hallucination in both alignment steps:\ supervised fine-tuning (SFT) and reinforcement learning (RL). In particular, we find that training the LLM on new knowledge or unfamiliar texts can encourage hallucination. This makes SFT less factual as it trains on human labeled data that may be novel to the LLM. Furthermore, reward functions used in standard RL can also encourage hallucination, because it guides the LLM to provide more helpful responses on a diverse set of instructions, often preferring longer and more detailed responses. Based on these observations, we propose factuality-aware alignment, comprised of factuality-aware SFT and factuality-aware RL through direct preference optimization. Experiments show that our proposed factuality-aware alignment guides LLMs to output more factual responses while maintaining instruction-following capability.