Alignment, bias, regulation, and responsible AI
![[2511.21331] The More, the Merrier: Contrastive Fusion for Higher-Order Multimodal Alignment](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
Abstract page for arXiv paper 2511.21331: The More, the Merrier: Contrastive Fusion for Higher-Order Multimodal Alignment
Abstract page for arXiv paper 2509.22367: What Is The Political Content in LLMs' Pre- and Post-Training Data?
Abstract page for arXiv paper 2507.22264: SmartCLIP: Modular Vision-language Alignment with Identification Guarantees
This paper presents a novel meta-cognitive architecture for AI in cybersecurity, advocating for a shift from traditional model-centric sy...
The paper presents TGIF, a novel approach to mitigate hallucinations in multimodal large language models (MLLMs) by leveraging a text-gui...
This article explores the intersection of active matter physics and robotics, focusing on the challenges faced by bio-inspired robotic sy...
This article evaluates the ability of small and reasoning large language models (LLMs) to assess journal article quality, revealing that ...
The paper presents MARS-Sep, a novel reinforcement learning framework for sound separation that enhances semantic consistency by aligning...
The paper presents a novel framework for premise verification in large language models (LLMs) to reduce hallucinations by using retrieval...
The article introduces the 'generative reasonable person,' a tool for assessing how ordinary people judge reasonableness in various legal...
The paper introduces PII-Bench, a novel framework for evaluating privacy protection systems in Large Language Models (LLMs), highlighting...
This article explores the differences between human-written and AI-generated texts by analyzing linguistic features, revealing significan...
The paper discusses the critical energy concerns associated with High-Performance Computing (HPC) systems and applications, emphasizing t...
This paper examines the shortcomings of speech recognition models in accurately transcribing high-stakes utterances, particularly U.S. st...
The paper introduces stable-worldmodel (SWM), a modular ecosystem for world modeling research that enhances reproducibility and standardi...
This article introduces ErrorMap and ErrorAtlas, innovative tools designed to analyze and categorize the failure patterns of large langua...
This paper explores the expressivity of structured argumentation frameworks that incorporate uncertainty, presenting both theoretical and...
MultiSHAP introduces a Shapley-based framework for explaining interactions in multimodal AI models, enhancing interpretability and trustw...
This paper presents a novel method for simulation-based inference that is robust to outliers and simplifies computation by eliminating th...
OpenAgentSafety introduces a modular framework for evaluating AI agent safety in real-world tasks, addressing critical vulnerabilities in...
This article explores the integration of Large Language Models (LLMs) with Uncrewed Aerial Vehicles (UAVs) for enhanced public safety, fo...
The paper introduces PolySHAP, an extension of KernelSHAP that uses interaction-informed polynomial regression to improve the accuracy of...
This article presents CARL-XRay, a novel continual learning framework for chest radiograph classification that adapts to new datasets wit...
Get the latest news, tools, and insights delivered to your inbox.
Daily or weekly digest • Unsubscribe anytime