We introduce a method for efficient multi-label text classification with large language models (LLMs), built on reformulating classification tasks as sequences of dichotomic (yes/no) decisions. Instead of generating all labels in a single structured response, each target dimension is queried independently, which, combined with a prefix caching mechanism, yields substantial efficiency gains for short-text inference without loss of accuracy. To demonstrate the approach, we focus on affective text analysis, covering 24 dimensions including emotions and sentiment. Using LLM-to-SLM distillation, a powerful annotator model (DeepSeek-V3) provides multiple annotations per text, which are aggregated to fine-tune smaller models (HerBERT-Large, CLARIN-1B, PLLuM-8B, Gemma3-1B). The fine-tuned models show significant improvements over zero-shot baselines, particularly on the dimensions seen during training. Our findings suggest that decomposing multi-label classification into dichotomic queries, combined with distillation and cache-aware inference, offers a scalable and effective framework for LLM-based classification. While we validate the method on affective states, the approach is general and applicable across domains.

The COVID-19 pandemic disrupted healthcare systems worldwide, disproportionately impacting individuals with chronic conditions such as cardiovascular disease (CVD). These disruptions -- through delayed care and behavioral changes, affected key CVD biomarkers, including LDL cholesterol (LDL-C), HbA1c, BMI, and systolic blood pressure (SysBP). Accurate modeling of these changes is crucial for predicting disease progression and guiding preventive care. However, prior work has not addressed multi-target prediction of CVD biomarker from Electronic Health Records (EHRs) using machine learning (ML), while jointly capturing biomarker interdependencies, temporal patterns, and predictive uncertainty. In this paper, we propose MBT-CB, a Multi-target Bayesian Transformer (MBT) with pre-trained BERT-based transformer framework to jointly predict LDL-C, HbA1c, BMI and SysBP CVD biomarkers from EHR data. The model leverages Bayesian Variational Inference to estimate uncertainties, embeddings to capture temporal relationships and a DeepMTR model to capture biomarker inter-relationships. We evaluate MBT-CT on retrospective EHR data from 3,390 CVD patient records (304 unique patients) in Central Massachusetts during the Covid-19 pandemic. MBT-CB outperformed a comprehensive set of baselines including other BERT-based ML models, achieving an MAE of 0.00887, RMSE of 0.0135 and MSE of 0.00027, while effectively capturing data and model uncertainty, patient biomarker inter-relationships, and temporal dynamics via its attention and embedding mechanisms. MBT-CB's superior performance highlights its potential to improve CVD biomarker prediction and support clinical decision-making during pandemics.

Large Language Models (LLMs) are rapidly being adopted by users across the globe, who interact with them in a diverse range of languages. At the same time, there are well-documented imbalances in the training data and optimisation objectives of this technology, raising doubts as to whether LLMs can represent the cultural diversity of their broad user base. In this study, we look at LLMs and cultural values and examine how prompt language and cultural framing influence model responses and their alignment with human values in different countries. We probe 10 LLMs with 63 items from the Hofstede Values Survey Module and World Values Survey, translated into 11 languages, and formulated as prompts with and without different explicit cultural perspectives. Our study confirms that both prompt language and cultural perspective produce variation in LLM outputs, but with an important caveat: While targeted prompting can, to a certain extent, steer LLM responses in the direction of the predominant values of the corresponding countries, it does not overcome the models' systematic bias toward the values associated with a restricted set of countries in our dataset: the Netherlands, Germany, the US, and Japan. All tested models, regardless of their origin, exhibit remarkably similar patterns: They produce fairly neutral responses on most topics, with selective progressive stances on issues such as social tolerance. Alignment with cultural values of human respondents is improved more with an explicit cultural perspective than with a targeted prompt language. Unexpectedly, combining both approaches is no more effective than cultural framing with an English prompt. These findings reveal that LLMs occupy an uncomfortable middle ground: They are responsive enough to changes in prompts to produce variation, but too firmly anchored to specific cultural defaults to adequately represent cultural diversity.

In this study, we experiment with the ability of LLMs to do temporal reasoning. Using a Norwegian book from 1940 containing trivia questions, we prompt the LLMs to answer the questions as if it were 1940. We also pose the questions in both English and Norwegian. Correct answers are often presented as sentences, and grading is done by means of LLM-as-judge, with sampled checks by a native speaker. Prompting in English consistently gave better results than in Norwegian, an unexpected result. In contrast, using larger LLMs improved results. We tested the DeepSeek-R1, Gemma3, Qwen3, and Llama3.1 model families, and also the largest available LLM especially crafted for Norwegian.

Recent advances in Large Language Models (LLMs) have highlighted the critical importance of extending context length, yet the quadratic complexity of attention mechanisms poses significant challenges for efficient long-context modeling. KV cache compression has emerged as a key approach to address this challenge. Through extensive empirical analysis, we reveal a fundamental yet previously overlooked asymmetry in KV caches: while adjacent keys receive similar attention weights ({\it local homogeneity}), adjacent values demonstrate distinct {\it heterogeneous} distributions. This key-value asymmetry reveals a critical limitation in existing compression methods that treat keys and values uniformly. To address the limitation, we propose a training-free compression framework (AsymKV) that combines homogeneity-based key merging with a mathematically proven lossless value compression. Extensive experiments demonstrate that AsymKV consistently outperforms existing long-context methods across various tasks and base models. For example, on LLaMA3.1-8B, AsymKV achieves an average score of 43.95 on LongBench, surpassing SOTA methods like H$_2$O (38.89) by a large margin.Our code can be found in this link:https://github.com/the-scale-lab/Asymkv.

Recent work suggests that large language models (LLMs) can perform multi-hop reasoning implicitly -- producing correct answers without explicitly verbalizing intermediate steps -- but the underlying mechanisms remain poorly understood. In this paper, we study how such implicit reasoning emerges by training transformers from scratch in a controlled symbolic environment. Our analysis reveals a three-stage developmental trajectory: early memorization, followed by in-distribution generalization, and eventually cross-distribution generalization. We find that training with atomic triples is not necessary but accelerates learning, and that second-hop generalization relies on query-level exposure to specific compositional structures. To interpret these behaviors, we introduce two diagnostic tools: cross-query semantic patching, which identifies semantically reusable intermediate representations, and a cosine-based representational lens, which reveals that successful reasoning correlates with the cosine-base clustering in hidden space. This clustering phenomenon in turn provides a coherent explanation for the behavioral dynamics observed across training, linking representational structure to reasoning capability. These findings provide new insights into the interpretability of implicit multi-hop reasoning in LLMs, helping to clarify how complex reasoning processes unfold internally and offering pathways to enhance the transparency of such models.

Abstract--T okenization is fundamental in assembly code analysis, impacting intrinsic characteristics like vocabulary size, semantic coverage, and extrinsic performance in downstream tasks. Despite its significance, tokenization in the context of assembly code remains an underexplored area. This study aims to address this gap by evaluating the intrinsic properties of Natural Language Processing (NLP) tokenization models and parameter choices, such as vocabulary size. We explore prepro-cessing customization options and pre-tokenization rules tailored to the unique characteristics of assembly code. Additionally, we assess their impact on downstream tasks like function signature prediction--a critical problem in binary code analysis. T o this end, we conduct a thorough study on various tokeniza-tion models, systematically analyzing their efficiency in encoding assembly instructions and capturing semantic nuances. Through intrinsic evaluations, we compare tokenizers based on tokeniza-tion efficiency, vocabulary compression, and representational fidelity for assembly code. Using state-of-the-art pre-trained models such as the decoder-only Large Language Model (LLM) Llama 3.2, the encoder-only transformer BERT, and the encoder-decoder model BART, we evaluate the effectiveness of these tokenizers across multiple performance metrics. Preliminary findings indicate that tokenizer choice significantly influences downstream performance, with intrinsic metrics providing partial but incomplete predictability of extrinsic evaluation outcomes. These results reveal complex trade-offs between intrinsic tokenizer properties and their utility in practical assembly code tasks. Ultimately, this study provides valuable insights into optimizing tokenization models for low-level code analysis, contributing to the robustness and scalability of Natural Language Model (NLM)-based binary analysis workflows. Tokenization is critical in transforming raw input data into structured representations, a process of utmost importance for Machine Learning (ML) and NLM model tasks [1]-[3]. While tokenization strategies have been studied extensively for natural [4] and high-level programming languages [5], assembly code presents unique challenges due to its low-level operations, diverse instruction sets, and non-standardized syntax across architectures. These challenges highlight the need for specialized tokenization techniques that effectively capture assembly code's structural and semantic intricacies [2]. Despite its importance, the role of tokenization in assembly code processing remains underexplored, particularly in its impact on downstream tasks involving modern NLMs. Recent research underscores the significant influence of tokenization on NLM model performance.

Large Language Models (LLMs) play a central role in modern artificial intelligence, yet their development has been primarily focused on English, resulting in limited support for other languages. We present PLLuM (Polish Large Language Model), the largest open-source family of foundation models tailored specifically for the Polish language. Developed by a consortium of major Polish research institutions, PLLuM addresses the need for high-quality, transparent, and culturally relevant language models beyond the English-centric commercial landscape. We describe the development process, including the construction of a new 140-billion-token Polish text corpus for pre-training, a 77k custom instructions dataset, and a 100k preference optimization dataset. A key component is a Responsible AI framework that incorporates strict data governance and a hybrid module for output correction and safety filtering. We detail the models' architecture, training procedures, and alignment techniques for both base and instruction-tuned variants, and demonstrate their utility in a downstream task within public administration. By releasing these models publicly, PLLuM aims to foster open research and strengthen sovereign AI technologies in Poland.

Adaptive batch size methods aim to accelerate neural network training, but existing approaches apply identical adaptation strategies across all architectures, assuming a one-size-fits-all solution. We introduce DEBA (Dynamic Efficient Batch Adaptation), an adaptive batch scheduler that monitors gradient variance, gradient norm variation and loss variation to guide batch size adaptations. Through systematic evaluation across six architectures (ResNet-18/50, DenseNet-121, EfficientNet-B0, MobileNet-V3, ViT-B16) on CIFAR-10 and CIFAR-100, with five random seeds per configuration, we demonstrate that the architecture fundamentally determines adaptation efficacy. Our findings reveal that: (1) lightweight and medium-depth architectures (MobileNet-V3, DenseNet-121, EfficientNet-B0) achieve a 45-62% training speedup with simultaneous accuracy improvements of 1-7%; (2) shallow residual networks (ResNet-18) show consistent gains of +2.4 - 4.0% in accuracy, 36 - 43% in speedup, while deep residual networks (ResNet-50) exhibit high variance and occasional degradation; (3) already-stable architectures (ViT-B16) show minimal speedup (6%) despite maintaining accuracy, indicating that adaptation benefits vary with baseline optimization characteristics. We introduce a baseline characterization framework using gradient stability metrics (stability score, gradient norm variation) that predicts which architectures will benefit from adaptive scheduling. Our ablation studies reveal critical design choices often overlooked in prior work: sliding window statistics (vs. full history) and sufficient cooldown periods (5+ epochs) between adaptations are essential for success. This work challenges the prevailing assumption that adaptive methods generalize across architectures and provides the first systematic evidence that batch size adaptation requires an architecture-aware design.

Reasoning language models perform well on complex tasks but are costly to deploy due to their size and long reasoning traces. We propose a routing approach that assigns each problem to the smallest model likely to solve it, reducing compute without sacrificing accuracy. Using intermediate representations from s1.1-32B, we train lightweight predictors of problem difficulty or model correctness to guide routing across a pool of reasoning models. On diverse math benchmarks, routing improves efficiency over random assignment and matches s1.1-32B's performance while using significantly less compute. Our results demonstrate that difficulty-aware routing is effective for cost-efficient deployment of reasoning models.