The rapid advancement of Large Language Models (LLMs) has spurred significant progress in Large Speech-Language Models (LSLMs), enhancing their capabilities in both speech understanding and generation. While existing LSLMs often concentrate on augmenting speech generation or tackling a diverse array of short-speech tasks, the efficient processing of long-form speech remains a critical yet underexplored challenge. This gap is primarily attributed to the scarcity of long-speech training datasets and the high computational costs associated with long sequences. To address these limitations, we introduce FastLongSpeech, a novel framework designed to extend LSLM capabilities for efficient long-speech processing without necessitating dedicated long-speech training data. FastLongSpeech incorporates an iterative fusion strategy that can compress excessively long-speech sequences into manageable lengths. To adapt LSLMs for long-speech inputs, it introduces a dynamic compression training approach, which exposes the model to short-speech sequences at varying compression ratios, thereby transferring the capabilities of LSLMs to long-speech tasks. To assess the long-speech capabilities of LSLMs, we develop a long-speech understanding benchmark called LongSpeech-Eval. Experiments show that our method exhibits strong performance in both long-speech and short-speech tasks, while greatly improving inference efficiency.

Although mathematics is often considered culturally neutral, the way mathematical problems are presented can carry implicit cultural context. Existing benchmarks like GSM8K are predominantly rooted in Western norms, including names, currencies, and everyday scenarios. In this work, we create culturally adapted variants of the GSM8K test set for five regions Africa, India, China, Korea, and Japan using prompt-based transformations followed by manual verification. We evaluate six large language models (LLMs), ranging from 8B to 72B parameters, across five prompting strategies to assess their robustness to cultural variation in math problem presentation. Our findings reveal a consistent performance gap: models perform best on the original US-centric dataset and comparatively worse on culturally adapted versions. However, models with reasoning capabilities are more resilient to these shifts, suggesting that deeper reasoning helps bridge cultural presentation gaps in mathematical tasks

Language models (LMs) are increasingly being deployed to perform autonomous data analyses. However, their data awareness -- the ability to recognize, reason over, and appropriately handle data artifacts such as missing values, outliers, and logical inconsistencies -- remains underexplored. These artifacts are especially common in real-world tabular data and, if mishandled, can significantly compromise the validity of analytical conclusions. To address this gap, we present RADAR, a benchmark for systematically evaluating data-aware reasoning on tabular data. We develop a framework to simulate data artifacts via programmatic perturbations to enable targeted evaluation of model behavior. RADAR comprises 2980 table query pairs, grounded in real-world data spanning 9 domains and 5 data artifact types. In addition to evaluating artifact handling, RADAR systematically varies table size to study how reasoning performance holds when increasing table size. Our evaluation reveals that, despite decent performance on tables without data artifacts, frontier models degrade significantly when data artifacts are introduced, exposing critical gaps in their capacity for robust, data-aware analysis. Designed to be flexible and extensible, RADAR supports diverse perturbation types and controllable table sizes, offering a valuable resource for advancing tabular reasoning.

Modern software systems increasingly integrate machine learning (ML) due to its advancements and ability to enhance data-driven decision-making. However, this integration introduces significant challenges for software engineering, especially in software product lines (SPLs), where managing variability and reuse becomes more complex with the inclusion of ML components. Although existing approaches have addressed variability management in SPLs and the integration of ML components in isolated systems, few have explored the intersection of both domains. Specifically, there is limited support for modeling and managing variability in SPLs that incorporate ML components. To bridge this gap, this article proposes a structured framework designed to extend Software Product Line engineering, facilitating the integration of ML components. It facilitates the design of SPLs with ML capabilities by enabling systematic modeling of variability and reuse. The proposal has been partially implemented with the VariaMos tool.

Supervised learning is classically formulated as training a model to minimize a fixed loss function over a fixed distribution, or task. However, an emerging paradigm instead views model training as extracting enough information from data so that the model can be used to minimize many losses on many downstream tasks. We formalize a mathematical framework for this paradigm, which we call panprediction, and study its statistical complexity. Formally, panprediction generalizes omniprediction and sits upstream from multi-group learning, which respectively focus on predictions that generalize to many downstream losses or many downstream tasks, but not both. Concretely, we design algorithms that learn deterministic and randomized panpredictors with $\tilde{O}(1/\varepsilon^3)$ and $\tilde{O}(1/\varepsilon^2)$ samples, respectively. Our results demonstrate that under mild assumptions, simultaneously minimizing infinitely many losses on infinitely many tasks can be as statistically easy as minimizing one loss on one task. Along the way, we improve the best known sample complexity guarantee of deterministic omniprediction by a factor of $1/\varepsilon$, and match all other known sample complexity guarantees of omniprediction and multi-group learning. Our key technical ingredient is a nearly lossless reduction from panprediction to a statistically efficient notion of calibration, called step calibration.

We present DialectalArabicMMLU, a new benchmark for evaluating the performance of large language models (LLMs) across Arabic dialects. While recently developed Arabic and multilingual benchmarks have advanced LLM evaluation for Modern Standard Arabic (MSA), dialectal varieties remain underrepresented despite their prevalence in everyday communication. DialectalArabicMMLU extends the MMLU-Redux framework through manual translation and adaptation of 3K multiple-choice question-answer pairs into five major dialects (Syrian, Egyptian, Emirati, Saudi, and Moroccan), yielding a total of 15K QA pairs across 32 academic and professional domains (22K QA pairs when also including English and MSA). The benchmark enables systematic assessment of LLM reasoning and comprehension beyond MSA, supporting both task-based and linguistic analysis. We evaluate 19 open-weight Arabic and multilingual LLMs (1B-13B parameters) and report substantial performance variation across dialects, revealing persistent gaps in dialectal generalization. DialectalArabicMMLU provides the first unified, human-curated resource for measuring dialectal understanding in Arabic, thus promoting more inclusive evaluation and future model development.

This paper presents Awal, a community-powered initiative for developing language technology resources for Tamazight. We provide a comprehensive review of the NLP landscape for Tamazight, examining recent progress in computational resources, and the emergence of community-driven approaches to address persistent data scarcity. Launched in 2024, awaldigital.org platform addresses the underrepresentation of Tamazight in digital spaces through a collaborative platform enabling speakers to contribute translation and voice data. We analyze 18 months of community engagement, revealing significant barriers to participation including limited confidence in written Tamazight and ongoing standardization challenges. Despite widespread positive reception, actual data contribution remained concentrated among linguists and activists. The modest scale of community contributions -- 6,421 translation pairs and 3 hours of speech data -- highlights the limitations of applying standard crowdsourcing approaches to languages with complex sociolinguistic contexts. We are working on improved open-source MT models using the collected data.

Spiking Neural Networks (SNNs) represent the latest generation of neural computation, offering a brain-inspired alternative to conventional Artificial Neural Networks (ANNs). Unlike ANNs, which depend on continuous-valued signals, SNNs operate using distinct spike events, making them inherently more energy-efficient and temporally dynamic. This study presents a comprehensive analysis of SNN design models, training algorithms, and multi-dimensional performance metrics, including accuracy, energy consumption, latency, spike count, and convergence behavior. Key neuron models such as the Leaky Integrate-and-Fire (LIF) and training strategies, including surrogate gradient descent, ANN-to-SNN conversion, and Spike-Timing Dependent Plasticity (STDP), are examined in depth. Results show that surrogate gradient-trained SNNs closely approximate ANN accuracy (within 1-2%), with faster convergence by the 20th epoch and latency as low as 10 milliseconds. Converted SNNs also achieve competitive performance but require higher spike counts and longer simulation windows. STDP-based SNNs, though slower to converge, exhibit the lowest spike counts and energy consumption (as low as 5 millijoules per inference), making them optimal for unsupervised and low-power tasks. These findings reinforce the suitability of SNNs for energy-constrained, latency-sensitive, and adaptive applications such as robotics, neuromorphic vision, and edge AI systems. While promising, challenges persist in hardware standardization and scalable training. This study concludes that SNNs, with further refinement, are poised to propel the next phase of neuromorphic computing.

Recommender systems (RSs) are intelligent filtering methods that suggest items to users based on their inferred preferences, derived from their interaction history on the platform. Collaborative filtering-based RSs rely on users' past interactions to generate recommendations. However, when a user is new to the platform--referred to as a cold-start user--there is no historical data available, making it difficult to provide personalized recommendations. To address this, rating elicitation techniques can be used to gather initial ratings or preferences on selected items, helping to build an early understanding of the user's tastes. Rating elicitation approaches are generally categorized into two types: non-personalized and personalized. Decision tree-based rating elicitation is a personalized method that queries users about their preferences at each node of the tree until sufficient information is gathered. In this paper, we propose an extension to the decision tree approach for rating elicitation in the context of music recommendation. Our method: (i) elicits not only item ratings but also preferences on attributes such as genres to better cluster users, and (ii) uses item pairs instead of single items at each node to more effectively learn user preferences. Experimental results demonstrate that both proposed enhancements lead to improved performance, particularly with a reduced number of queries.

Evaluating the abilities of large language models (LLMs) for tasks that require long-term memory and thus long-context reasoning, for example in conversational settings, is hampered by the existing benchmarks, which often lack narrative coherence, cover narrow domains, and only test simple recall-oriented tasks. This paper introduces a comprehensive solution to these challenges. First, we present a novel framework for automatically generating long (up to 10M tokens), coherent, and topically diverse conversations, accompanied by probing questions targeting a wide range of memory abilities. From this, we construct BEAM, a new benchmark comprising 100 conversations and 2,000 validated questions. Second, to enhance model performance, we propose LIGHT-a framework inspired by human cognition that equips LLMs with three complementary memory systems: a long-term episodic memory, a short-term working memory, and a scratchpad for accumulating salient facts. Our experiments on BEAM reveal that even LLMs with 1M token context windows (with and without retrieval-augmentation) struggle as dialogues lengthen. In contrast, LIGHT consistently improves performance across various models, achieving an average improvement of 3.5%-12.69% over the strongest baselines, depending on the backbone LLM. An ablation study further confirms the contribution of each memory component.