Meta

Multi-Objective Alignment aims to align Large Language Models (LLMs) with diverse and often conflicting human values by optimizing multiple objectives simultaneously. Existing methods predominantly rely on static preference weight construction strategies. However, rigidly aligning to fixed targets discards valuable intermediate information, as training responses inherently embody valid preference trade-offs even when deviating from the target. To address this limitation, we propose Meal, i.e., M

We propose Mochi, a Graph Foundation Model that addresses task unification and training efficiency by adopting a meta-learning based training framework. Prior models pre-train with reconstruction-based objectives such as link prediction, and assume that the resulting representations can be aligned with downstream tasks through a separate unification step such as class prototypes. We demonstrate through synthetic and real-world experiments that this procedure, while simple and intuitive, has limi

We present Six Llamas, a comparative study examining whether large language models fine-tuned on distinct religious corpora encode systematically different patterns of ethical reasoning. Six variants of Meta-Llama-3.1-8B are constructed: one unmodified control and five LoRA-adapted models trained exclusively on the sacred and theological texts of Christianity, Islam, Judaism, Hinduism, or Buddhism. All six models are probed with an identical battery of 17 standardized ethical prompts spanning mo

Existing AI agent safety benchmarks focus on generic criminal harm (cybercrime, harassment, weapon synthesis), leaving a systematic blind spot for a distinct and commercially consequential threat category: agents harming their own deployers. Real-world incidents illustrate the gap: Slack AI credential exfiltration (Aug 2024), Microsoft 365 Copilot calendar-injection leaks (Jan 2024), and a Meta agent unauthorized forum post exposing operational data (Mar 2026). We propose Owner-Harm, a formal th

Bayesian optimization (BO) has for sequential optimization of expensive black-box functions demonstrated practicality and effectiveness in many real-world settings. Meta-Bayesian optimization (meta-BO) focuses on improving the sample efficiency of BO by making use of information from related tasks. Although meta-BO is sample-efficient when task structure transfers, poor alignment between meta-training and test tasks can cause suboptimal queries to be suggested during online optimization. To this

The essence of current neural network architectures is phenomenological fitting: they learn input-output statistical correlations via massive parameters and data, yet lack intrinsic understanding of the fundamental principles governing physical reality. This paper proposes a paradigm leap from pure phenomenological fitting to the fusion of phenomenological fitting and endogenous deduction. By embedding physical meta-principles into neural network architecture, we construct the Meta-Principle Phy

This work introduces a modular platform that brings together six AI services, automatic speech recognition via OpenAI Whisper, multilingual translation through Meta NLLB, speech synthesis using AWS Polly, emotion classification with RoBERTa, dialogue summarisation via flan t5 base samsum, and International Sign (IS) rendering through Google MediaPipe. A corpus of IS gesture recordings was processed to derive hand landmark coordinates, which were subsequently mapped onto three dimensional avatar

Clinical evidence synthesis requires identifying relevant trials from large registries and aggregating results that account for population differences. While recent LLM-based approaches have automated components of systematic review, they do not support end-to-end evidence synthesis. Moreover, conventional meta-analysis weights studies by statistical precision without considering clinical compatibility reflected in eligibility criteria. We propose EligMeta, an agentic framework that integrates a

Automated radiology report generation from 3D computed tomography (CT) volumes is challenging due to extreme sequence lengths, severe class imbalance, and the tendency of large language models (LLMs) to ignore visual tokens in favor of linguistic priors. We present Ker-VLJEPA-3B, a four-phase curriculum learning framework for free-text report generation from thoracic CT volumes. A phased training curriculum progressively adapts a Llama 3.2 3B decoder to ground its output in visual features from

EngGPT2-16B-A3B is the latest iteration of Engineering Group's Italian LLM and it's built to be a Sovereign, Efficient and Open model. EngGPT2 is trained on 2.5 trillion tokens - less than Qwen3's 36T or Llama3's 15T - and delivers performance on key benchmarks, including MMLU-Pro, GSM8K, IFEval and HumanEval, comparable to dense models in the 8B-16B range, while requiring one-fifth to half of the inference power, and between one-tenth to one-sixth of the training data and consequent needed trai

Existing test-time scaling (TTS) methods for unified multimodal models (UMMs) in text-to-image (T2I) generation primarily rely on search or sampling strategies that produce only instance-level improvements, limiting the ability to learn from prior inferences and accumulate knowledge across similar prompts. To overcome these limitations, we propose Meta-TTRL, a metacognitive test-time reinforcement learning framework. Meta-TTRL performs test-time parameter optimization guided by model-intrinsic m

Computer-Use Agents (CUAs) are emerging as a new paradigm in human-computer interaction, enabling autonomous execution of tasks in desktop environment by perceiving high-level natural-language instructions. As such agents become increasingly capable and are deployed across diverse desktop environments, evaluating their behavior in a scalable and reliable manner becomes a critical challenge. Existing evaluation pipelines rely on static benchmarks, rule-based success checks, or manual inspection,

Alignment safety research assumes that ethical instructions improve model behavior, but how language models internally process such instructions remains unknown. We conducted over 600 multi-agent simulations across four models (Llama 3.3 70B, GPT-4o mini, Qwen3-Next-80B-A3B, Sonnet 4.5), four ethical instruction formats (none, minimal norm, reasoned norm, virtue framing), and two languages (Japanese, English). Confirmatory analysis fully replicated the Llama Japanese dissociation pattern from a

Language models are commonly fine-tuned for safety alignment to refuse harmful prompts. One approach fine-tunes them to generate categorical refusal tokens that distinguish different refusal types before responding. In this work, we leverage a version of Llama 3 8B fine-tuned with these categorical refusal tokens to enable inference-time control over fine-grained refusal behavior, improving both safety and reliability. We show that refusal token fine-tuning induces separable, category-aligned di

Large Language Models (LLMs) are increasingly being deployed in multilingual, multicultural settings, yet their reliance on predominantly English-centric training data risks misalignment with the diverse cultural values of different societies. In this paper, we present a comprehensive, multilingual audit of the cultural alignment of contemporary LLMs including GPT-4o-Mini, Gemini-2.5-Flash, Llama 3.2, Mistral and Gemma 3 across India, East Asia and Southeast Asia. Our study specifically focuses

Millions of patients are regularly using large language model (LLM) chatbots for medical advice, raising patient safety concerns. This physician-led red-teaming study compares the safety of four publicly available chatbots-Claude by Anthropic, Gemini by Google, GPT-4o by OpenAI, and Llama-3.0/3.1-70B by Meta-on a new dataset, HealthAdvice, using an evaluation framework that enables quantitative and qualitative analysis. In total, 888 chatbot responses are evaluated for 222 patient-posed advice-s

Abstract Conventional electromagnetic cloaking paradigms predominantly necessitate the encasing of static objects within predefined topological enclosures, fundamentally restricting invisibility to fixed, closed geometries. Realizing dynamic, adaptive concealment for arbitrary moving targets within an open, boundary-free aperture remains a formidable challenge. Here, we report a meta-reinforcement-learning metasurface (Meta 2 Surface) that enables the first experimental demonstration of a "trans

BACKGROUND: With the continuous advancement of cancer treatment technology, the proportion of cancer survivors is gradually increasing, but they also face many psychological challenges. These challenges can seriously affect their quality of life. Telemedicine, as an innovative medical service model, can be combined with psychosocial intervention to provide cancer survivors with convenient, economical and accessible services to assist them in more effectively managing the difficulties posed by ca

Background: Theintegration of artificial intelligence (AI) into clinical decision support systems (CDSSs) has significantly enhanced diagnostic precision, risk stratification, and treatment planning. AI models remain a barrier to clinical adoption, emphasizing the critical role of explainable AI (XAI). Methods: This systematic meta-analysis synthesizes findings from 62 peer-reviewed studies published between 2018 and 2025, examining the use of XAI methods within CDSSs across various clinical dom

Large Language Models (LLMs) hold promise for medical applications but often lack domain-specific expertise. Retrieval Augmented Generation (RAG) enables customization by integrating specialized knowledge. This study assessed the accuracy, consistency, and safety of LLM-RAG models in determining surgical fitness and delivering preoperative instructions using 35 local and 23 international guidelines. Ten LLMs (e.g., GPT3.5, GPT4, GPT4o, Gemini, Llama2, and Llama3, Claude) were tested across 14 cl

INTRODUCTION: The integration of artificial intelligence in mental health care represents a transformative shift in the identification, treatment, and management of mental disorders. This systematic review explores the diverse applications of artificial intelligence, emphasizing both its benefits and associated challenges. METHODS: A comprehensive literature search was conducted across multiple databases based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses, including ProQu

Modern education faces persistent challenges, including disengagement, inequitable access to learning resources, and the lack of personalized instruction, particularly in virtual environments. In this perspective, we envision a transformative Metaverse classroom model, the Multi-layered Immersive Learning Environment (Meta-MILE) to address these critical issues. The Meta-MILE framework integrates essential components such as immersive infrastructure, personalized interactions, social collaborati

As Large Language Models (LLMs) become increasingly integrated into secure software development workflows, a critical question remains unanswered: can these models not only detect insecure code but also reliably classify vulnerabilities according to standardized taxonomies? In this work, we conduct a systematic evaluation of three state-of-the-art LLMs - Llama3, Codestral, and Deepseek R1 - using a carefully filtered subset of the Big-Vul dataset annotated with eight representative Common Weakne

This paper argues that a range of current AI systems have learned how to deceive humans. We define deception as the systematic inducement of false beliefs in the pursuit of some outcome other than the truth. We first survey empirical examples of AI deception, discussing both special-use AI systems (including Meta's CICERO) and general-purpose AI systems (including large language models). Next, we detail several risks from AI deception, such as fraud, election tampering, and losing control of AI.

Artificial intelligence (AI) has emerged as a transformative force in various sectors, including medicine and healthcare. Large language models like ChatGPT showcase AI's potential by generating human-like text through prompts. ChatGPT's adaptability holds promise for reshaping medical practices, improving patient care, and enhancing interactions among healthcare professionals, patients, and data. In pandemic management, ChatGPT rapidly disseminates vital information. It serves as a virtual assi

Since Facebook was renamed Meta, a lot of attention, debate, and exploration have intensified about what the Metaverse is, how it works, and the possible ways to exploit it. It is anticipated that Metaverse will be a continuum of rapidly emerging technologies, usecases, capabilities, and experiences that will make it up for the next evolution of the Internet. Several researchers have already surveyed the literature on artificial intelligence (AI) and wireless communications in realizing the Meta

Objective The primary aim of this research was to address the limitations observed in the medical knowledge of prevalent large language models (LLMs) such as ChatGPT, by creating a specialized language model with enhanced accuracy in medical advice. Methods We achieved this by adapting and refining the large language model meta-AI (LLaMA) using a large dataset of 100,000 patient-doctor dialogues sourced from a widely used online medical consultation platform. These conversations were cleaned and

Abstract The initial hype and fanfare from the Meta Platforms view of how the metaverse could be brought to life has evolved into an ongoing discussion of not only the metaverse's impact on users and organizations but also the societal and cultural implications of widespread usage. The potential of consumer interaction with brands within the metaverse has engendered significant debate within the marketing‐focused discourse on the key challenges and transformative opportunities for marketers. Dra

Data infrastructures, economic processes, and governance models of digital platforms are increasingly pervading urban sectors and spheres of urban life. This phenomenon is known as platformization, which has in turn given rise to the phenomena of platform society, where platforms have permeated the core of urban societies. A recent manifestation of platformization is the Metaverse, a global platform project launched by Meta (formerly Facebook) as a globally operating platform company. The Metave

Abstract This paper addresses quantitative meta-analyses on corporate governance-related determinants and firms’ (non) financial consequences of Corporate Social Responsibility (CSR). Legitimacy theory as our theoretical framework assumes that, through a social contract, a company must fulfil the respective society’s values and expectations and gain legitimacy. We also rely on the business case argument, assuming a positive relationship between CSR and financial outcomes of the firm. This analys

BACKGROUND: Global efforts toward the development and deployment of a vaccine for COVID-19 are rapidly advancing. To achieve herd immunity, widespread administration of vaccines is required, which necessitates significant cooperation from the general public. As such, it is crucial that governments and public health agencies understand public sentiments toward vaccines, which can help guide educational campaigns and other targeted policy interventions. OBJECTIVE: The aim of this study was to deve

The construction industry is known to be overwhelmed with resource planning, risk management and logistic challenges which often result in design defects, project delivery delays, cost overruns and contractual disputes. These challenges have instigated research in the application of advanced machine learning algorithms such as deep learning to help with diagnostic and prescriptive analysis of causes and preventive measures. However, the publicity created by tech firms like Google, Facebook and A

An accessible synthesis of ethical issues raised by artificial intelligence that moves beyond hype and nightmare scenarios to address concrete questions. Artificial intelligence powers Google's search engine, enables Facebook to target advertising, and allows Alexa and Siri to do their jobs. AI is also behind self-driving cars, predictive policing, and autonomous weapons that can kill without human intervention. These and other AI applications raise complex ethical issues that are the subject of

With the exponential growth in the mobile device market over the last decade, chatbots are becoming an increasingly popular option to interact with users, and their popularity and adoption are rapidly spreading. These mobile devices change the way we communicate and allow ever-present learning in various environments. This study examined educational chatbots for Facebook Messenger to support learning. The independent web directory was screened to assess chatbots for this study resulting in the i

BACKGROUND: Eating disorders among adolescents are an ongoing public health concern. Sustainable health promotion programmes require a thorough understanding of the social context in which minors engage. Initial studies show that young people make extensive use of social networks in order to exchange experiences and gather information. During this process their (buying) behaviour is significantly affected by so-called influencers. METHODS: The exploratory research studies non-campaign driven hea

Background: Frequent expression of negative emotion words on social media has been linked to depression. However, metrics have relied on average values, not dynamic measures of emotional volatility. Objective: The aim of this study was to report on the associations between depression severity and the variability (time-unstructured) and instability (time-structured) in emotion word expression on Facebook and Twitter across status updates. Methods: Status updates and depression severity ratings of

Purpose The purpose of this paper is to explore the utilization of the social network, Facebook, as an instrument of stakeholder engagement and dialogic accounting in American charitable foundations, specifically non-profit organizations that are dedicated to philanthropy. Design/methodology/approach The research motivation involves whether online interaction through Facebook could represent a channel of dialogic accounting that engages organizational stakeholders. This paper aims to understand

Abstract Natural language processing (NLP) techniques can be used to make inferences about peoples’ mental states from what they write on Facebook, Twitter and other social media. These inferences can then be used to create online pathways to direct people to health information and assistance and also to generate personalized interventions. Regrettably, the computational methods used to collect, process and utilize online writing data, as well as the evaluations of these techniques, are still di

Purpose – The purpose of this paper is to assess if online interaction through social media, particularly Facebook, Twitter, and YouTube, represents an effective stakeholder engagement mechanism in order to define the contents of social, environmental, or sustainability reporting (SESR). Design/methodology/approach – After examining 332 worldwide sustainability reports for the year 2013, drawn up according to the guidelines provided by the Global Reporting Initiative, the authors conducted a con

New technologies like large-scale social media sites (e.g., Facebook and Twitter) and crowdsourcing services (e.g., Amazon Mechanical Turk, Crowdflower, Clickworker) are impacting social science research and providing many new and interesting avenues for research. The use of these new technologies for research has not been without challenges, and a recently published psychological study on Facebook has led to a widespread discussion of the ethics of conducting large-scale experiments online. Sur