Hakim Hacid

Wenbin Li, Hakim Hacid, Ebtesam Almazrouei et al.

The union of Edge Computing (EC) and Artificial Intelligence (AI) has brought forward the Edge AI concept to provide intelligent solutions close to the end-user environment, for privacy preservation, low latency to real-time performance, and resource optimization. Machine Learning (ML), as the most advanced branch of AI in the past few years, has shown encouraging results and applications in the edge environment. Nevertheless, edge-powered ML solutions are more complex to realize due to the joint constraints from both edge computing and AI domains, and the corresponding solutions are expected to be efficient and adapted in technologies such as data processing, model compression, distributed inference, and advanced learning paradigms for Edge ML requirements. Despite the fact that a great deal of the attention garnered by Edge ML is gained in both the academic and industrial communities, we noticed the lack of a complete survey on existing Edge ML technologies to provide a common understanding of this concept. To tackle this, this paper aims at providing a comprehensive taxonomy and a systematic review of Edge ML techniques, focusing on the soft computing aspects of existing paradigms and techniques. We start by identifying the Edge ML requirements driven by the joint constraints. We then extensively survey more than twenty paradigms and techniques along with their representative work, covering two main parts: edge inference, and edge learning. In particular, we analyze how each technique fits into Edge ML by meeting a subset of the identified requirements. We also summarize Edge ML frameworks and open issues to shed light on future directions for Edge ML.

Quentin Malartic, Nilabhra Roy Chowdhury, Ruxandra Cojocaru et al.

We introduce Falcon2-11B, a foundation model trained on over five trillion tokens, and its multimodal counterpart, Falcon2-11B-vlm, which is a vision-to-text model. We report our findings during the training of the Falcon2-11B which follows a multi-stage approach where the early stages are distinguished by their context length and a final stage where we use a curated, high-quality dataset. Additionally, we report the effect of doubling the batch size mid-training and how training loss spikes are affected by the learning rate. The downstream performance of the foundation model is evaluated on established benchmarks, including multilingual and code datasets. The foundation model shows strong generalization across all the tasks which makes it suitable for downstream finetuning use cases. For the vision language model, we report the performance on several benchmarks and show that our model achieves a higher average score compared to open-source models of similar size. The model weights and code of both Falcon2-11B and Falcon2-11B-vlm are made available under a permissive license.

Talal Algumaei, Ruben Solozabal, Reda Alami et al.

This work studies non-cooperative Multi-Agent Reinforcement Learning (MARL) where multiple agents interact in the same environment and whose goal is to maximize the individual returns. Challenges arise when scaling up the number of agents due to the resultant non-stationarity that the many agents introduce. In order to address this issue, Mean Field Games (MFG) rely on the symmetry and homogeneity assumptions to approximate games with very large populations. Recently, deep Reinforcement Learning has been used to scale MFG to games with larger number of states. Current methods rely on smoothing techniques such as averaging the q-values or the updates on the mean-field distribution. This work presents a different approach to stabilize the learning based on proximal updates on the mean-field policy. We name our algorithm Mean Field Proximal Policy Optimization (MF-PPO), and we empirically show the effectiveness of our method in the OpenSpiel framework.

Jingwei Zuo, Wenbin Li, Michele Baldo et al.

Air quality forecasting has garnered significant attention recently, with data-driven models taking center stage due to advancements in machine learning and deep learning models. However, researchers face challenges with complex data acquisition and the lack of open-sourced datasets, hindering efficient model validation. This paper introduces PurpleAirSF, a comprehensive and easily accessible dataset collected from the PurpleAir network. With its high temporal resolution, various air quality measures, and diverse geographical coverage, this dataset serves as a useful tool for researchers aiming to develop novel forecasting models, study air pollution patterns, and investigate their impacts on health and the environment. We present a detailed account of the data collection and processing methods employed to build PurpleAirSF. Furthermore, we conduct preliminary experiments using both classic and modern spatio-temporal forecasting models, thereby establishing a benchmark for future air quality forecasting tasks.

Vitor Gaboardi dos Santos, Ibrahim Khadraoui, Ibrahim Farhat et al.

Large Language Models (LLMs) have recently empowered agentic frameworks to exhibit advanced reasoning and planning capabilities. However, their integration in robotic control pipelines remains limited in two aspects: (1) prior \ac{llm}-based approaches often lack modular, agentic execution mechanisms, limiting their ability to plan, reflect on outcomes, and revise actions in a closed-loop manner; and (2) existing benchmarks for manipulation tasks focus on low-level control and do not systematically evaluate multistep reasoning and linguistic variation. In this paper, we propose Agentic LLM for Robot Manipulation (ALRM), an LLM-driven agentic framework for robotic manipulation. ALRM integrates policy generation with agentic execution through a ReAct-style reasoning loop, supporting two complementary modes: Code-asPolicy (CaP) for direct executable control code generation, and Tool-as-Policy (TaP) for iterative planning and tool-based action execution. To enable systematic evaluation, we also introduce a novel simulation benchmark comprising 56 tasks across multiple environments, capturing linguistically diverse instructions. Experiments with ten LLMs demonstrate that ALRM provides a scalable, interpretable, and modular approach for bridging natural language reasoning with reliable robotic execution. Results reveal Claude-4.1-Opus as the top closed-source model and Falcon-H1-7B as the top open-source model under CaP.

Jingwei Zuo, George Arvanitakis, Hakim Hacid

Human activity recognition (HAR) has been a classic research problem. In particular, with recent machine learning (ML) techniques, the recognition task has been largely investigated by companies and integrated into their products for customers. However, most of them apply a predefined activity set and conduct the learning process on the cloud, hindering specific personalizations from end users (i.e., edge devices). Even though recent progress in Incremental Learning allows learning new-class data on the fly, the learning process is generally conducted on the cloud, requiring constant data exchange between cloud and edge devices, thus leading to data privacy issues. In this paper, we propose PILOTE, which pushes the incremental learning process to the extreme edge, while providing reliable data privacy and practical utility, e.g., low processing latency, personalization, etc. In particular, we consider the practical challenge of extremely limited data during the incremental learning process on edge, where catastrophic forgetting is required to be handled in a practical way. We validate PILOTE with extensive experiments on human activity data collected from mobile sensors. The results show PILOTE can work on edge devices with extremely limited resources while providing reliable performance.

Parteek Jamwal, Minghao Shao, Boyuan Chen et al.

Large Language Models (LLMs) have demonstrated remarkable capabilities across various cybersecurity tasks, including vulnerability classification, detection, and patching. However, their potential in automated vulnerability report documentation and analysis remains underexplored. We present RAVEN (Retrieval Augmented Vulnerability Exploration Network), a framework leveraging LLM agents and Retrieval Augmented Generation (RAG) to synthesize comprehensive vulnerability analysis reports. Given vulnerable source code, RAVEN generates reports following the Google Project Zero Root Cause Analysis template. The framework uses four modules: an Explorer agent for vulnerability identification, a RAG engine retrieving relevant knowledge from curated databases including Google Project Zero reports and CWE entries, an Analyst agent for impact and exploitation assessment, and a Reporter agent for structured report generation. To ensure quality, RAVEN includes a task specific LLM Judge evaluating reports across structural integrity, ground truth alignment, code reasoning quality, and remediation quality. We evaluate RAVEN on 105 vulnerable code samples covering 15 CWE types from the NIST-SARD dataset. Results show an average quality score of 54.21%, supporting the effectiveness of our approach for automated vulnerability documentation.

Reda Alami, Ali Khalifa Almansoori, Ahmed Alzubaidi et al.

We demonstrate that preference optimization methods can effectively enhance LLM safety. Applying various alignment techniques to the Falcon 11B model using safety datasets, we achieve a significant boost in global safety score (from $57.64\%$ to $99.90\%$) as measured by LlamaGuard 3 8B, competing with state-of-the-art models. On toxicity benchmarks, average scores in adversarial settings dropped from over $0.6$ to less than $0.07$. However, this safety improvement comes at the cost of reduced general capabilities, particularly in math, suggesting a trade-off. We identify noise contrastive alignment (Safe-NCA) as an optimal method for balancing safety and performance. Our study ultimately shows that alignment techniques can be sufficient for building safe and robust models.

Jingwei Zuo, Wenbin Li, Michele Baldo et al.

Air Quality Monitoring and Forecasting has been a popular research topic in recent years. Recently, data-driven approaches for air quality forecasting have garnered significant attention, owing to the availability of well-established data collection facilities in urban areas. Fixed infrastructures, typically deployed by national institutes or tech giants, often fall short in meeting the requirements of diverse personalized scenarios, e.g., forecasting in areas without any existing infrastructure. Consequently, smaller institutes or companies with limited budgets are compelled to seek tailored solutions by introducing more flexible infrastructures for data collection. In this paper, we propose an expandable graph attention network (EGAT) model, which digests data collected from existing and newly-added infrastructures, with different spatial structures. Additionally, our proposal can be embedded into any air quality forecasting models, to apply to the scenarios with evolving spatial structures. The proposal is validated over real air quality data from PurpleAir.

Jingwei Zuo, Maksim Velikanov, Ilyas Chahed et al.

In this report, we introduce Falcon-H1, a new series of large language models (LLMs) featuring hybrid architecture designs optimized for both high performance and efficiency across diverse use cases. Unlike earlier Falcon models built solely on Transformer or Mamba architectures, Falcon-H1 adopts a parallel hybrid approach that combines Transformer-based attention with State Space Models (SSMs), known for superior long-context memory and computational efficiency. We systematically revisited model design, data strategy, and training dynamics, challenging conventional practices in the field. Falcon-H1 is released in multiple configurations, including base and instruction-tuned variants at 0.5B, 1.5B, 1.5B-deep, 3B, 7B, and 34B parameters. Quantized instruction-tuned models are also available, totaling over 30 checkpoints on Hugging Face Hub. Falcon-H1 models demonstrate state-of-the-art performance and exceptional parameter and training efficiency. The flagship Falcon-H1-34B matches or outperforms models up to 70B scale, such as Qwen3-32B, Qwen2.5-72B, and Llama3.3-70B, while using fewer parameters and less data. Smaller models show similar trends: the Falcon-H1-1.5B-Deep rivals current leading 7B-10B models, and Falcon-H1-0.5B performs comparably to typical 7B models from 2024. These models excel across reasoning, mathematics, multilingual tasks, instruction following, and scientific knowledge. With support for up to 256K context tokens and 18 languages, Falcon-H1 is suitable for a wide range of applications. All models are released under a permissive open-source license, underscoring our commitment to accessible and impactful AI research.

Sofian Chaybouti, Sanath Narayan, Yasser Dahou et al.

Vision foundation models trained via multi-teacher distillation offer a promising path toward unified visual representations, yet the learning dynamics and data efficiency of such approaches remain underexplored. In this paper, we systematically study multi-teacher distillation for vision foundation models and identify key factors that enable training at lower computational cost. We introduce Agglomerative Mixture-of-Experts Vision Foundation Models (AMoE), which distill knowledge from SigLIP2 and DINOv3 simultaneously into a Mixture-of-Experts student. We show that (1) our Asymmetric Relation-Knowledge Distillation loss preserves the geometric properties of each teacher while enabling effective knowledge transfer, (2) token-balanced batching that packs varying-resolution images into sequences with uniform token budgets stabilizes representation learning across resolutions without sacrificing performance, and (3) hierarchical clustering and sampling of training data--typically reserved for self-supervised learning--substantially improves sample efficiency over random sampling for multi-teacher distillation. By combining these findings, we curate OpenLVD200M, a 200M-image corpus that demonstrates superior efficiency for multi-teacher distillation. Instantiated in a Mixture-of-Experts. We release OpenLVD200M and distilled models.

Falcon LLM Team, Iheb Chaabane, Puneesh Khanna et al.

This work introduces Falcon-H1R, a 7B-parameter reasoning-optimized model that establishes the feasibility of achieving competitive reasoning performance with small language models (SLMs). Falcon-H1R stands out for its parameter efficiency, consistently matching or outperforming SOTA reasoning models that are $2\times$ to $7\times$ larger across a variety of reasoning-intensive benchmarks. These results underscore the importance of careful data curation and targeted training strategies (via both efficient SFT and RL scaling) in delivering significant performance gains without increasing model size. Furthermore, Falcon-H1R advances the 3D limits of reasoning efficiency by combining faster inference (through its hybrid-parallel architecture design), token efficiency, and higher accuracy. This unique blend makes Falcon-H1R-7B a practical backbone for scaling advanced reasoning systems, particularly in scenarios requiring extensive chain-of-thoughts generation and parallel test-time scaling. Leveraging the recently introduced DeepConf approach, Falcon-H1R achieves state-of-the-art test-time scaling efficiency, offering substantial improvements in both accuracy and computational cost. As a result, Falcon-H1R demonstrates that compact models, through targeted model training and architectural choices, can deliver robust and scalable reasoning performance.

Adam Younsi, Ahmed Attia, Abdalgader Abubaker et al.

Achieving both accuracy and diverse reasoning remains challenging for Large Language Models (LLMs) in complex domains like mathematics. A key bottleneck is evaluating intermediate reasoning steps to guide generation without costly human annotations. To address this, we first introduce a novel Process Reward Model (PRM) trained automatically using Monte Carlo Tree Search coupled with a similarity-based data augmentation technique, effectively capturing step-level reasoning quality. Leveraging this PRM, we then adapt Generative Flow Networks (GFlowNets) to operate at the reasoning step level. Unlike traditional reinforcement learning focused on maximizing a single reward, GFlowNets naturally sample diverse, high-quality solutions proportional to their rewards, as measured by our PRM. Empirical evaluation shows strong improvements in both accuracy and solution diversity on challenging mathematical benchmarks (e.g., +2.59% absolute accuracy on MATH Level 5 for Llama3.2-3B), with effective generalization to unseen datasets (+9.4\% absolute on SAT MATH). Furthermore, we benchmark our PRM against existing open-source reward models, demonstrating superior alignment with reasoning quality and more consistent guidance for downstream generation. Our work demonstrates the potential of PRM-guided, step-level GFlowNets for developing more robust and versatile mathematical reasoning in LLMs.

Maksim Velikanov, Ilyas Chahed, Jingwei Zuo et al.

Applying weight decay (WD) to matrix layers is standard practice in large-language-model pretraining. Prior work suggests that stochastic gradient noise induces a Brownian-like expansion of the weight matrices W, whose growth is counteracted by WD, leading to a WD-noise equilibrium with a certain weight norm ||W||. In this work, we view the equilibrium norm as a harmful artifact of the training procedure, and address it by introducing learnable multipliers to learn the optimal scale. First, we attach a learnable scalar multiplier to W and confirm that the WD-noise equilibrium norm is suboptimal: the learned scale adapts to data and improves performance. We then argue that individual row and column norms are similarly constrained, and free their scale by introducing learnable per-row and per-column multipliers. Our method can be viewed as a learnable, more expressive generalization of muP multipliers. It outperforms a well-tuned muP baseline, reduces the computational overhead of multiplier tuning, and surfaces practical questions such as forward-pass symmetries and the width-scaling of the learned multipliers. Finally, we validate learnable multipliers with both Adam and Muon optimizers, where it shows improvement in downstream evaluations matching the improvement of the switching from Adam to Muon.

George Arvanitakis, Jingwei Zuo, Mthandazo Ndhlovu et al.

Edge Machine Learning (Edge ML), which shifts computational intelligence from cloud-based systems to edge devices, is attracting significant interest due to its evident benefits including reduced latency, enhanced data privacy, and decreased connectivity reliance. While these advantages are compelling, they introduce unique challenges absent in traditional cloud-based approaches. In this paper, we delve into the intricacies of Edge-based learning, examining the interdependencies among: (i) constrained data storage on Edge devices, (ii) limited computational power for training, and (iii) the number of learning classes. Through experiments conducted using our MAGNETO system, that focused on learning human activities via data collected from mobile sensors, we highlight these challenges and offer valuable perspectives on Edge ML.

Aviraj Bevli, Sofian Chaybouti, Yasser Dahou et al.

Perception-centric systems are typically implemented with a modular encoder-decoder pipeline: a vision backbone for feature extraction and a separate decoder (or late-fusion module) for task prediction. This raises a central question: is this architectural separation essential or can a single early-fusion stack do both perception and task modeling at scale? We introduce Falcon Perception, a unified dense Transformer that processes image patches and text tokens in a shared parameter space from the first layer, using a hybrid attention pattern (bidirectional among image tokens, causal for prediction tokens) to combine global visual context with autoregressive, variable-length instance generation. To keep dense outputs practical, Falcon Perception retains a lightweight token interface and decodes continuous spatial outputs with specialized heads, enabling parallel high-resolution mask prediction. Our design promotes simplicity: we keep a single scalable backbone and shift complexity toward data and training signals, adding only small heads where outputs are continuous and dense. On SA-Co, Falcon Perception improves mask quality to 68.0 Macro-F$_1$ compared to 62.3 of SAM3. We also introduce PBench, a benchmark targeting compositional prompts (OCR, spatial constraints, relations) and dense long-context regimes, where the model shows better gains. Finally, we extend the same early-fusion recipe to Falcon OCR: a compact 300M-parameter model which attains 80.3% on olmOCR and 88.64 on OmniDocBench.

Leen AlQadi, Ahmed Alzubaidi, Mohammed Alyafeai et al.

We present QIMMA, a quality-assured Arabic LLM leaderboard that places systematic benchmark validation at its core. Rather than aggregating existing resources as-is, QIMMA applies a multi-model assessment pipeline combining automated LLM judgment with human review to surface and resolve systematic quality issues in well-established Arabic benchmarks before evaluation. The result is a curated, multi-domain, multi-task evaluation suite of over 52k samples, grounded predominantly in native Arabic content; code evaluation tasks are the sole exception, as they are inherently language-agnostic. Transparent implementation via LightEval, EvalPlus and public release of per-sample inference outputs make QIMMA a reproducible and community-extensible foundation for Arabic NLP evaluation.

Gokul Karthik Kumar, Ludovick Lepauloux, Hakim Hacid

Whisper has become the de-facto encoder for extracting general-purpose audio features in large audio-language models, where a 30-second clip is typically represented by 1500 frame features projected into an LLM. In contrast, audio-text embedding models like CLAP-based models have largely relied on alternative audio encoders (e.g., HTS-AT, PaSST), and have not leveraged Whisper effectively. We present WavLink, a compact audio-text embedding model that augments Whisper encoder with a learnable global token, trained jointly with a text encoder. Through a systematic study of design choices, including pretrained text encoders, loss functions, training modes, and data mixtures, we identify configurations that yield state-of-the-art retrieval performance. Our two-stage training recipe across three model sizes, combined with Matryoshka-style supervision, improves scalability, enabling 8x smaller embeddings with minimal performance drop. WavLink also demonstrates competitive performance on AIR-Bench with MCQs and zero-shot classification.

Aymen Rayane Khouas, Mohamed Reda Bouadjenek, Hakim Hacid et al.

Edge computing has gained significant traction in recent years, promising enhanced efficiency by integrating artificial intelligence capabilities at the edge. While the focus has primarily been on the deployment and inference of Machine Learning (ML) models at the edge, the training aspect remains less explored. This survey, explores the concept of edge learning, specifically the optimization of ML model training at the edge. The objective is to comprehensively explore diverse approaches and methodologies in edge learning, synthesize existing knowledge, identify challenges, and highlight future trends. Utilizing Scopus and Web of science advanced search, relevant literature on edge learning was identified, revealing a concentration of research efforts in distributed learning methods, particularly federated learning. This survey further provides a guideline for comparing techniques used to optimize ML for edge learning, along with an exploration of the different frameworks, libraries, and simulation tools available. In doing so, the paper contributes to a holistic understanding of the current landscape and future directions in the intersection of edge computing and machine learning, paving the way for informed comparisons between optimization methods and techniques designed for training on the edge.

Ngoc Dung Huynh, Mohamed Reda Bouadjenek, Sunil Aryal et al.

Visual Question Answering (VQA) is an evolving research field aimed at enabling machines to answer questions about visual content by integrating image and language processing techniques such as feature extraction, object detection, text embedding, natural language understanding, and language generation. With the growth of multimodal data research, VQA has gained significant attention due to its broad applications, including interactive educational tools, medical image diagnosis, customer service, entertainment, and social media captioning. Additionally, VQA plays a vital role in assisting visually impaired individuals by generating descriptive content from images. This survey introduces a taxonomy of VQA architectures, categorizing them based on design choices and key components to facilitate comparative analysis and evaluation. We review major VQA approaches, focusing on deep learning-based methods, and explore the emerging field of Large Visual Language Models (LVLMs) that have demonstrated success in multimodal tasks like VQA. The paper further examines available datasets and evaluation metrics essential for measuring VQA system performance, followed by an exploration of real-world VQA applications. Finally, we highlight ongoing challenges and future directions in VQA research, presenting open questions and potential areas for further development. This survey serves as a comprehensive resource for researchers and practitioners interested in the latest advancements and future

Aymane El Firdoussi, Mohamed El Amine Seddik, Soufiane Hayou et al.

Synthetic data has gained attention for training large language models, but poor-quality data can harm performance (see, e.g., Shumailov et al. (2023); Seddik et al. (2024)). A potential solution is data pruning, which retains only high-quality data based on a score function (human or machine feedback). Previous work Feng et al. (2024) analyzed models trained on synthetic data as sample size increases. We extend this by using random matrix theory to derive the performance of a binary classifier trained on a mix of real and pruned synthetic data in a high dimensional setting. Our findings identify conditions where synthetic data could improve performance, focusing on the quality of the generative model and verification strategy. We also show a smooth phase transition in synthetic label noise, contrasting with prior sharp behavior in infinite sample limits. Experiments with toy models and large language models validate our theoretical results.

Shijing Chen, Mohamed Reda Bouadjenek, Shoaib Jameel et al.

Multi-level Hierarchical Classification (MLHC) tackles the challenge of categorizing items within a complex, multi-layered class structure. However, traditional MLHC classifiers often rely on a backbone model with independent output layers, which tend to ignore the hierarchical relationships between classes. This oversight can lead to inconsistent predictions that violate the underlying taxonomy. Leveraging Large Language Models (LLMs), we propose a novel taxonomy-embedded transitional LLM-agnostic framework for multimodality classification. The cornerstone of this advancement is the ability of models to enforce consistency across hierarchical levels. Our evaluations on the MEP-3M dataset - a multi-modal e-commerce product dataset with various hierarchical levels - demonstrated a significant performance improvement compared to conventional LLM structures.

Ruiyi Yang, Hao Xue, Imran Razzak et al.

Graph Retrieval-Augmented Generation (GraphRAG) has proven highly effective in enhancing the performance of Large Language Models (LLMs) on tasks that require external knowledge. By leveraging Knowledge Graphs (KGs), GraphRAG improves information retrieval for complex reasoning tasks, providing more precise and comprehensive retrieval and generating more accurate responses to QAs. However, most RAG methods fall short in addressing multi-step reasoning, particularly when both information extraction and inference are necessary. To address this limitation, this paper presents Knowledge Graph-Based Iterative Retrieval-Augmented Generation (KG-IRAG), a novel framework that integrates KGs with iterative reasoning to improve LLMs' ability to handle queries involving temporal and logical dependencies. Through iterative retrieval steps, KG-IRAG incrementally gathers relevant data from external KGs, enabling step-by-step reasoning. The proposed approach is particularly suited for scenarios where reasoning is required alongside dynamic temporal data extraction, such as determining optimal travel times based on weather conditions or traffic patterns. Experimental results show that KG-IRAG improves accuracy in complex reasoning tasks by effectively integrating external knowledge with iterative, logic-based retrieval. Additionally, three new datasets: weatherQA-Irish, weatherQA-Sydney, and trafficQA-TFNSW, are formed to evaluate KG-IRAG's performance, demonstrating its potential beyond traditional RAG applications.

Jingwei Zuo, George Arvanitakis, Mthandazo Ndhlovu et al.

Human activity recognition (HAR) is a well-established field, significantly advanced by modern machine learning (ML) techniques. While companies have successfully integrated HAR into consumer products, they typically rely on a predefined activity set, which limits personalizations at the user level (edge devices). Despite advancements in Incremental Learning for updating models with new data, this often occurs on the Cloud, necessitating regular data transfers between cloud and edge devices, thus leading to data privacy issues. In this paper, we propose MAGNETO, an Edge AI platform that pushes HAR tasks from the Cloud to the Edge. MAGNETO allows incremental human activity learning directly on the Edge devices, without any data exchange with the Cloud. This enables strong privacy guarantees, low processing latency, and a high degree of personalization for users. In particular, we demonstrate MAGNETO in an Android device, validating the whole pipeline from data collection to result visualization.

Basma El Amel Boussaha, Leen AlQadi, Mugariya Farooq et al.

Arabic is one of the most widely spoken languages in the world, yet efforts to develop and evaluate Large Language Models (LLMs) for Arabic remain relatively limited. Most existing Arabic benchmarks focus on linguistic, cultural, or religious content, leaving a significant gap in domains like STEM and code which are increasingly relevant for real-world LLM applications. To help bridge this gap, we present 3LM, a suite of three benchmarks designed specifically for Arabic. The first is a set of STEM-related question-answer pairs, naturally sourced from Arabic textbooks and educational worksheets. The second consists of synthetically generated STEM questions, created using the same sources. The third benchmark focuses on code generation, built through a careful translation of two widely used code benchmarks, incorporating a human-in-the-loop process with several rounds of review to ensure high-quality and faithful translations. We release all three benchmarks publicly to support the growth of Arabic LLM research in these essential but underrepresented areas.

Ruiyi Yang, Hao Xue, Imran Razzak et al.

Retrieval-Augmented Generation (RAG) systems empower large language models (LLMs) with external knowledge, yet struggle with efficiency-accuracy trade-offs when scaling to large knowledge graphs. Existing approaches often rely on monolithic graph retrieval, incurring unnecessary latency for simple queries and fragmented reasoning for complex multi-hop questions. To address these challenges, this paper propose SPLIT-RAG, a multi-agent RAG framework that addresses these limitations with question-driven semantic graph partitioning and collaborative subgraph retrieval. The innovative framework first create Semantic Partitioning of Linked Information, then use the Type-Specialized knowledge base to achieve Multi-Agent RAG. The attribute-aware graph segmentation manages to divide knowledge graphs into semantically coherent subgraphs, ensuring subgraphs align with different query types, while lightweight LLM agents are assigned to partitioned subgraphs, and only relevant partitions are activated during retrieval, thus reduce search space while enhancing efficiency. Finally, a hierarchical merging module resolves inconsistencies across subgraph-derived answers through logical verifications. Extensive experimental validation demonstrates considerable improvements compared to existing approaches.

Ubaid Azam, Imran Razzak, Shelly Vishwakarma et al.

AI-driven medical predictions with trustworthy confidence are essential for ensuring the responsible use of AI in healthcare applications. The growing capabilities of AI raise questions about their trustworthiness in healthcare, particularly due to opaque decision-making and limited data availability. This paper proposes a novel approach to address these challenges, introducing a Bayesian Monte Carlo Dropout model with kernel modelling. Our model is designed to enhance reliability on small medical datasets, a crucial barrier to the wider adoption of AI in healthcare. This model leverages existing language models for improved effectiveness and seamlessly integrates with current workflows. Extensive evaluations of public medical datasets showcase our model's superior performance across diverse tasks. We demonstrate significant improvements in reliability, even with limited data, offering a promising step towards building trust in AI-driven medical predictions and unlocking its potential to improve patient care.

Ruiyi Yang, Hao Xue, Imran Razzak et al.

Retrieval-augmented generation (RAG) remains brittle on multi-step questions and heterogeneous evidence sources, trading accuracy against latency and token/tool budgets. This paper introducesHierarchical Sequence (HSEQ) Iteration for Heterogeneous Question Answering, a unified framework that (i) linearize documents, tables, and knowledge graphs into a reversible hierarchical sequence with lightweight structural tags, and (ii) perform structure-aware iteration to collect just-enough evidence before answer synthesis. A Head Agent provides guidance that leads retrieval, while an Iteration Agent selects and expands HSeq via structure-respecting actions (e.g., parent/child hops, table row/column neighbors, KG relations); Finally the head agent composes canonicalized evidence to genearte the final answer, with an optional refinement loop to resolve detected contradictions. Experiments on HotpotQA (text), HybridQA/TAT-QA (table+text), and MetaQA (KG) show consistent EM/F1 gains over strong single-pass, multi-hop, and agentic RAG baselines with high efficiency. Besides, HSEQ exhibits three key advantages: (1) a format-agnostic unification that enables a single policy to operate across text, tables, and KGs without per-dataset specialization; (2) guided, budget-aware iteration that reduces unnecessary hops, tool calls, and tokens while preserving accuracy; and (3) evidence canonicalization for reliable QA, improving answers consistency and auditability.

Ahmed Alzubaidi, Shaikha Alsuwaidi, Basma El Amel Boussaha et al.

This survey provides the first systematic review of Arabic LLM benchmarks, analyzing 40+ evaluation benchmarks across NLP tasks, knowledge domains, cultural understanding, and specialized capabilities. We propose a taxonomy organizing benchmarks into four categories: Knowledge, NLP Tasks, Culture and Dialects, and Target-Specific evaluations. Our analysis reveals significant progress in benchmark diversity while identifying critical gaps: limited temporal evaluation, insufficient multi-turn dialogue assessment, and cultural misalignment in translated datasets. We examine three primary approaches: native collection, translation, and synthetic generation discussing their trade-offs regarding authenticity, scale, and cost. This work serves as a comprehensive reference for Arabic NLP researchers, providing insights into benchmark methodologies, reproducibility standards, and evaluation metrics while offering recommendations for future development.

Gokul Karthik Kumar, Rishabh Saraf, Ludovick Lepauloux et al.

Large language models (LLMs) have transformed NLP, yet their integration with audio remains underexplored -- despite audio's centrality to human communication. We introduce Falcon3-Audio, a family of Audio-Language Models (ALMs) built on instruction-tuned LLMs and Whisper encoders. Using a remarkably small amount of public audio data -- less than 30K hours (5K unique) -- Falcon3-Audio-7B matches the best reported performance among open-weight models on the MMAU benchmark, with a score of 64.14, matching R1-AQA, while distinguishing itself through superior data and parameter efficiency, single-stage training, and transparency. Notably, our smallest 1B model remains competitive with larger open models ranging from 2B to 13B parameters. Through extensive ablations, we find that common complexities -- such as curriculum learning, multiple audio encoders, and intricate cross-attention connectors -- are not required for strong performance, even compared to models trained on over 500K hours of data.

Aymen Rayane Khouas, Mohamed Reda Bouadjenek, Hakim Hacid et al.

Graph federated recommendation systems offer a privacy-preserving alternative to traditional centralized recommendation architectures, which often raise concerns about data security. While federated learning enables personalized recommendations without exposing raw user data, existing aggregation methods overlook the unique properties of user embeddings in this setting. Indeed, traditional aggregation methods fail to account for their complexity and the critical role of user similarity in recommendation effectiveness. Moreover, evolving user interactions require adaptive aggregation while preserving the influence of high-relevance anchor users (the primary users before expansion in graph-based frameworks). To address these limitations, we introduce Dist-FedAvg, a novel distance-based aggregation method designed to enhance personalization and aggregation efficiency in graph federated learning. Our method assigns higher aggregation weights to users with similar embeddings, while ensuring that anchor users retain significant influence in local updates. Empirical evaluations on multiple datasets demonstrate that Dist-FedAvg consistently outperforms baseline aggregation techniques, improving recommendation accuracy while maintaining seamless integration into existing federated learning frameworks.

Mouadh Yagoubi, Yasser Dahou, Billel Mokeddem et al.

Existing benchmarks have proven effective for assessing the performance of fully trained large language models. However, we find striking differences in the early training stages of small models, where benchmarks often fail to provide meaningful or discriminative signals. To explore how these differences arise, this competition tackles the challenge of designing scientific knowledge evaluation tasks specifically tailored for measuring early training progress of language models. Participants are invited to develop novel evaluation methodologies or adapt existing benchmarks to better capture performance differences among language models. To support this effort, we provide three pre-trained small models (0.5B, 1B, and 3B parameters), along with intermediate checkpoints sampled during training up to 200B tokens. All experiments and development work can be run on widely available free cloud-based GPU platforms, making participation accessible to researchers with limited computational resources. Submissions will be evaluated based on three criteria: the quality of the performance signal they produce, the consistency of model rankings at 1 trillion tokens of training, and their relevance to the scientific knowledge domain. By promoting the design of tailored evaluation strategies for early training, this competition aims to attract a broad range of participants from various disciplines, including those who may not be machine learning experts or have access to dedicated GPU resources. Ultimately, this initiative seeks to make foundational LLM research more systematic and benchmark-informed from the earliest phases of model development.

Mohammed Djameleddine Belgoumri, Mohamed Reda Bouadjenek, Hakim Hacid et al.

Training large neural networks (NNs) requires optimizing high-dimensional data-dependent loss functions. The optimization landscape of these functions is often highly complex and textured, even fractal-like, with many spurious local minima, ill-conditioned valleys, degenerate points, and saddle points. Complicating things further is the fact that these landscape characteristics are a function of the data, meaning that noise in the training data can propagate forward and give rise to unrepresentative small-scale geometry. This poses a difficulty for gradient-based optimization methods, which rely on local geometry to compute updates and are, therefore, vulnerable to being derailed by noisy data. In practice,this translates to a strong dependence of the optimization dynamics on the noise in the data, i.e., poor generalization performance. To remediate this problem, we propose a new optimization procedure: Rolling Ball Optimizer (RBO), that breaks this spatial locality by incorporating information from a larger region of the loss landscape in its updates. We achieve this by simulating the motion of a rigid sphere of finite radius rolling on the loss landscape, a straightforward generalization of Gradient Descent (GD) that simplifies into it in the infinitesimal limit. The radius serves as a hyperparameter that determines the scale at which RBO sees the loss landscape, allowing control over the granularity of its interaction therewith. We are motivated by the intuition that the large-scale geometry of the loss landscape is less data-specific than its fine-grained structure, and that it is easier to optimize. We support this intuition by proving that our algorithm has a smoothing effect on the loss function. Evaluation against SGD, SAM, and Entropy-SGD, on MNIST and CIFAR-10/100 demonstrates promising results in terms of convergence speed, training accuracy, and generalization performance.

Shijing Chen, Shoaib Jameel, Mohamed Reda Bouadjenek et al.

Traditional Multi-level Hierarchical Classification (MLHC) classifiers often rely on backbone models with $n$ independent output layers. This structure tends to overlook the hierarchical relationships between classes, leading to inconsistent predictions that violate the underlying taxonomy. Additionally, once a backbone architecture for an MLHC classifier is selected, adapting the model to accommodate new tasks can be challenging. For example, incorporating fairness to protect sensitive attributes within a hierarchical classifier necessitates complex adjustments to maintain the class hierarchy while enforcing fairness constraints. In this paper, we extend this concept to hierarchical classification by introducing a fair, model-agnostic layer designed to enforce taxonomy and optimize specific objectives, including consistency, fairness, and exact match. Our evaluations demonstrate that the proposed layer not only improves the fairness of predictions but also enforces the taxonomy, resulting in consistent predictions and superior performance. Compared to Large Language Models (LLMs) employing in-processing de-biasing techniques and models without any bias correction, our approach achieves better outcomes in both fairness and accuracy, making it particularly valuable in sectors like e-commerce, healthcare, and education, where predictive reliability is crucial.

Ghadir Alselwi, Hao Xue, Shoaib Jameel et al.

Effective long-term memory management is crucial for language models handling extended contexts. We introduce a novel framework that dynamically ranks memory entries based on relevance. Unlike previous works, our model introduces a novel relevance scoring and a pointwise re-ranking model for key-value embeddings, inspired by learning-to-rank techniques in information retrieval. Enhanced Ranked Memory Augmented Retrieval ERMAR achieves state-of-the-art results on standard benchmarks.

Ngoc Dung Huynh, Mohamed Reda Bouadjenek, Sunil Aryal et al.

Visual Question Answering (VQA) has emerged as a promising area of research to develop AI-based systems for enabling interactive and immersive learning. Numerous VQA datasets have been introduced to facilitate various tasks, such as answering questions or identifying unanswerable ones. However, most of these datasets are constructed using real-world images, leaving the performance of existing models on cartoon images largely unexplored. Hence, in this paper, we present "SimpsonsVQA", a novel dataset for VQA derived from The Simpsons TV show, designed to promote inquiry-based learning. Our dataset is specifically designed to address not only the traditional VQA task but also to identify irrelevant questions related to images, as well as the reverse scenario where a user provides an answer to a question that the system must evaluate (e.g., as correct, incorrect, or ambiguous). It aims to cater to various visual applications, harnessing the visual content of "The Simpsons" to create engaging and informative interactive systems. SimpsonsVQA contains approximately 23K images, 166K QA pairs, and 500K judgments (https://simpsonsvqa.org). Our experiments show that current large vision-language models like ChatGPT4o underperform in zero-shot settings across all three tasks, highlighting the dataset's value for improving model performance on cartoon images. We anticipate that SimpsonsVQA will inspire further research, innovation, and advancements in inquiry-based learning VQA.

Salem Lahlou, Abdalgader Abubaker, Hakim Hacid

Preference optimization methods have been successfully applied to improve not only the alignment of large language models (LLMs) with human values, but also specific natural language tasks such as summarization and stylistic continuations. This paper proposes using preference optimization methods on Chain-of-Thought steps in order to improve the mathematical reasoning performances of language models. While the chosen answers are obtained from datasets that include reasoning traces, we propose two complementary schemes for generating rejected answers: weak LLM prompting, and digit corruption. Our approach leads to increased accuracy on the GSM8K and AQuA-RAT mathematical reasoning benchmarks for Falcon2-11B and Mistral-7B. Additionally, the improved abilities transfer to non-mathematical tasks, including the ARC benchmark and symbolic reasoning challenges. For example, our method can lead to up to relative 8.47% and 18.73% increases in accuracy on the GSM8K and AQuA benchmarks respectively, without any extra annotations. This work suggests that the path towards better language reasoning abilities goes through spending resources on creating high-quality datasets of reasoning traces.

Mohammed Djameleddine Belgoumri, Mohamed Reda Bouadjenek, Sunil Aryal et al.

Data-driven Artificial Intelligence (AI) systems trained using Machine Learning (ML) are shaping an ever-increasing (in size and importance) portion of our lives, including, but not limited to, recommendation systems, autonomous driving technologies, healthcare diagnostics, financial services, and personalized marketing. On the one hand, the outsized influence of these systems imposes a high standard of quality, particularly in the data used to train them. On the other hand, establishing and maintaining standards of Data Quality (DQ) becomes more challenging due to the proliferation of Edge Computing and Internet of Things devices, along with their increasing adoption for training and deploying ML models. The nature of the edge environment -- characterized by limited resources, decentralized data storage, and processing -- exacerbates data-related issues, making them more frequent, severe, and difficult to detect and mitigate. From these observations, it follows that DQ research for edge ML is a critical and urgent exploration track for the safety and robust usefulness of present and future AI systems. Despite this fact, DQ research for edge ML is still in its infancy. The literature on this subject remains fragmented and scattered across different research communities, with no comprehensive survey to date. Hence, this paper aims to fill this gap by providing a global view of the existing literature from multiple disciplines that can be grouped under the umbrella of DQ for edge ML. Specifically, we present a tentative definition of data quality in Edge computing, which we use to establish a set of DQ dimensions. We explore each dimension in detail, including existing solutions for mitigation.

Ubaid Azam, Imran Razzak, Shelly Vishwakarma et al.

Predicting legal judgments with reliable confidence is paramount for responsible legal AI applications. While transformer-based deep neural networks (DNNs) like BERT have demonstrated promise in legal tasks, accurately assessing their prediction confidence remains crucial. We present a novel Bayesian approach called BayesJudge that harnesses the synergy between deep learning and deep Gaussian Processes to quantify uncertainty through Bayesian kernel Monte Carlo dropout. Our method leverages informative priors and flexible data modelling via kernels, surpassing existing methods in both predictive accuracy and confidence estimation as indicated through brier score. Extensive evaluations of public legal datasets showcase our model's superior performance across diverse tasks. We also introduce an optimal solution to automate the scrutiny of unreliable predictions, resulting in a significant increase in the accuracy of the model's predictions by up to 27\%. By empowering judges and legal professionals with more reliable information, our work paves the way for trustworthy and transparent legal AI applications that facilitate informed decisions grounded in both knowledge and quantified uncertainty.

Jingwei Zuo, Hakim Hacid

Human Activity Recognition (HAR) has been studied for decades, from data collection, learning models, to post-processing and result interpretations. However, the inherent hierarchy in the activities remains relatively under-explored, despite its significant impact on model performance and interpretation. In this paper, we propose H-HAR, by rethinking the HAR tasks from a fresh perspective by delving into their intricate global label relationships. Rather than building multiple classifiers separately for multi-layered activities, we explore the efficacy of a flat model enhanced with graph-based label relationship modeling. Being hierarchy-aware, the graph-based label modeling enhances the fundamental HAR model, by incorporating intricate label relationships into the model. We validate the proposal with a multi-label classifier on complex human activity data. The results highlight the advantages of the proposal, which can be vertically integrated into advanced HAR models to further enhance their performances.