Large Language Models (LLMs) are primarily trained on high-resource natural languages, limiting their effectiveness in low-resource settings and in tasks requiring deep logical reasoning. This research introduces Rosetta-PL, a benchmark designed to evaluate LLMs' logical reasoning and generalization capabilities in a controlled environment. We construct Rosetta-PL by translating a dataset of logical propositions from Lean into a custom logical language, which is then used to fine-tune an LLM (e.g., GPT-4o). Our experiments analyze the impact of the size of the dataset and the translation methodology on the performance of the model. Our results indicate that preserving logical relationships in the translation process significantly boosts precision, with accuracy plateauing beyond roughly 20,000 training samples. These insights provide valuable guidelines for optimizing LLM training in formal reasoning tasks and improving performance in various low-resource language applications.

Advancements in generative Artificial Intelligence (AI) hold great promise for automating radiology workflows, yet challenges in interpretability and reliability hinder clinical adoption. This paper presents an automated radiology report generation framework that combines Concept Bottleneck Models (CBMs) with a Multi-Agent Retrieval-Augmented Generation (RAG) system to bridge AI performance with clinical explainability. CBMs map chest X-ray features to human-understandable clinical concepts, enabling transparent disease classification. Meanwhile, the RAG system integrates multi-agent collaboration and external knowledge to produce contextually rich, evidence-based reports. Our demonstration showcases the system's ability to deliver interpretable predictions, mitigate hallucinations, and generate high-quality, tailored reports with an interactive interface addressing accuracy, trust, and usability challenges. This framework provides a pathway to improving diagnostic consistency and empowering radiologists with actionable insights.

Telegram has become a major space for political discourse and alternative media. However, its lack of moderation allows misinformation, extremism, and toxicity to spread. While prior research focused on these particular phenomena or topics, these have mostly been examined separately, and a broader understanding of the Telegram ecosystem is still missing. In this work, we fill this gap by conducting a large-scale analysis of the Italian Telegram sphere, leveraging a dataset of 186 million messages from 13,151 chats collected in 2023. Using network analysis, Large Language Models, and toxicity detection tools, we examine how different thematic communities form, align ideologically, and engage in harmful discourse within the Italian cultural context. Results show strong thematic and ideological homophily. We also identify mixed ideological communities where far-left and far-right rhetoric coexist on particular geopolitical issues. Beyond political analysis, we find that toxicity, rather than being isolated in a few extreme chats, appears widely normalized within highly toxic communities. Moreover, we find that Italian discourse primarily targets Black people, Jews, and gay individuals independently of the topic. Finally, we uncover common trend of intra-national hostility, where Italians often attack other Italians, reflecting regional and intra-regional cultural conflicts that can be traced back to old historical divisions. This study provides the first large-scale mapping of the Italian Telegram ecosystem, offering insights into ideological interactions, toxicity, and identity-targets of hate and contributing to research on online toxicity across different cultural and linguistic contexts on Telegram.

This paper explores how cloud-native databases enable AI-driven applications by leveraging purpose-built technologies such as vector databases (pgvector), graph databases (AWS Neptune), NoSQL stores (Amazon DocumentDB, DynamoDB), and relational cloud databases (Aurora MySQL and PostgreSQL). It presents architectural patterns for integrating AI workloads with cloud databases, including Retrieval-Augmented Generation (RAG) [1] with LLMs, real-time data pipelines, AI-driven query optimization, and embeddings-based search. Performance benchmarks, scalability considerations, and cost-efficient strategies are evaluated to guide the design of AI-enabled applications. Real-world case studies from industries such as healthcare, finance, and customer experience illustrate how enterprises utilize cloud databases to enhance AI capabilities while ensuring security, governance, and compliance with enterprise and regulatory standards. By providing a comprehensive analysis of AI and cloud database integration, this paper serves as a practical guide for researchers, architects, and enterprises to build next-generation AI applications that optimize performance, scalability, and cost efficiency in cloud environments.

Large Language Models (LLMs) are increasingly leveraged for translation tasks but often fall short when translating inclusive language -- such as texts containing the singular 'they' pronoun or otherwise reflecting fair linguistic protocols. Because these challenges span both computational and societal domains, it is imperative to critically evaluate how well LLMs handle inclusive translation with a well-founded framework. This paper presents FairTranslate, a novel, fully human-annotated dataset designed to evaluate non-binary gender biases in machine translation systems from English to French. FairTranslate includes 2418 English-French sentence pairs related to occupations, annotated with rich metadata such as the stereotypical alignment of the occupation, grammatical gender indicator ambiguity, and the ground-truth gender label (male, female, or inclusive). We evaluate four leading LLMs (Gemma2-2B, Mistral-7B, Llama3.1-8B, Llama3.3-70B) on this dataset under different prompting procedures. Our results reveal substantial biases in gender representation across LLMs, highlighting persistent challenges in achieving equitable outcomes in machine translation. These findings underscore the need for focused strategies and interventions aimed at ensuring fair and inclusive language usage in LLM-based translation systems. We make the FairTranslate dataset publicly available on Hugging Face, and disclose the code for all experiments on GitHub.

Code-generating Large Language Models (LLMs) have become essential tools in modern software development, enhancing productivity and accelerating development. This paper aims to investigate the fine-tuning of code-generating LLMs using Reinforcement Learning and Direct Preference Optimization, further improving their performance. To achieve this, we enhance the training data for the reward model with the help of symbolic execution techniques, ensuring more comprehensive and objective data. With symbolic execution, we create a custom dataset that better captures the nuances in code evaluation. Our reward models, fine-tuned on this dataset, demonstrate significant improvements over the baseline, CodeRL, in estimating the quality of generated code. Our code-generating LLMs, trained with the help of reward model feedback, achieve similar results compared to the CodeRL benchmark.

Understanding the inner workings of neural embeddings, particularly in models such as BERT, remains a challenge because of their high-dimensional and opaque nature. This paper proposes a framework for uncovering the specific dimensions of vector embeddings that encode distinct linguistic properties (LPs). We introduce the Linguistically Distinct Sentence Pairs (LDSP-10) dataset, which isolates ten key linguistic features such as synonymy, negation, tense, and quantity. Using this dataset, we analyze BERT embeddings with various methods, including the Wilcoxon signed-rank test, mutual information, and recursive feature elimination, to identify the most influential dimensions for each LP. We introduce a new metric, the Embedding Dimension Impact (EDI) score, which quantifies the relevance of each embedding dimension to a LP. Our findings show that certain properties, such as negation and polarity, are robustly encoded in specific dimensions, while others, like synonymy, exhibit more complex patterns. This study provides insights into the interpretability of embeddings, which can guide the development of more transparent and optimized language models, with implications for model bias mitigation and the responsible deployment of AI systems.

Recent advancements in language multimodal models (LMMs) for video have demonstrated their potential for understanding video content, yet the task of comprehending multi-discipline lectures remains largely unexplored. We introduce Video-MMLU, a massive benchmark designed to evaluate the capabilities of LMMs in understanding Multi-Discipline Lectures. We evaluate over 90 open-source and proprietary models, ranging from 0.5B to 40B parameters. Our results highlight the limitations of current models in addressing the cognitive challenges presented by these lectures, especially in tasks requiring both perception and reasoning. Additionally, we explore how the number of visual tokens and the large language models influence performance, offering insights into the interplay between multimodal perception and reasoning in lecture comprehension.

This technical report presents a cost-efficient strategy for training a video generation foundation model. We present a mid-sized research model with approximately 7 billion parameters (7B) called Seaweed-7B trained from scratch using 665,000 H100 GPU hours. Despite being trained with moderate computational resources, Seaweed-7B demonstrates highly competitive performance compared to contemporary video generation models of much larger size. Design choices are especially crucial in a resource-constrained setting. This technical report highlights the key design decisions that enhance the performance of the medium-sized diffusion model. Empirically, we make two observations: (1) Seaweed-7B achieves performance comparable to, or even surpasses, larger models trained on substantially greater GPU resources, and (2) our model, which exhibits strong generalization ability, can be effectively adapted across a wide range of downstream applications either by lightweight fine-tuning or continue training. See the project page at https://seaweed.video/

Large language models (LLMs) often produce inaccurate or misleading content-hallucinations. To address this challenge, we introduce Noise-Augmented Fine-Tuning (NoiseFiT), a novel framework that leverages adaptive noise injection based on the signal-to-noise ratio (SNR) to enhance model robustness. In particular, NoiseFiT selectively perturbs layers identified as either high-SNR (more robust) or low-SNR (potentially under-regularized) using a dynamically scaled Gaussian noise. We further propose a hybrid loss that combines standard cross-entropy, soft cross-entropy, and consistency regularization to ensure stable and accurate outputs under noisy training conditions. Our theoretical analysis shows that adaptive noise injection is both unbiased and variance-preserving, providing strong guarantees for convergence in expectation. Empirical results on multiple test and benchmark datasets demonstrate that NoiseFiT significantly reduces hallucination rates, often improving or matching baseline performance in key tasks. These findings highlight the promise of noise-driven strategies for achieving robust, trustworthy language modeling without incurring prohibitive computational overhead. Given the comprehensive and detailed nature of our experiments, we have publicly released the fine-tuning logs, benchmark evaluation artifacts, and source code online at W&B, Hugging Face, and GitHub, respectively, to foster further research, accessibility and reproducibility.