arXiv subjects

Wolfgang Nejdl

Publications and source records attributed to Wolfgang Nejdl.

At least 19 recordsLinked to original sources

arXiv2026

Expanding the Lexicon of Ge'ez Based African Languages: A Comparative Study of Amharic and Tigrinya

Multilingual pre-trained language models such as XLM-R perform well for major languages but struggle with low-resource Ge'ez-script languages, largely due to high out-of-vocabulary (OOV) rates and excessive subword fragmentation from Latin-script-centric tokenizers. We introduce VEXMLM, a vocabulary-extended variant of XLM-R targeting Amharic and Tigrinya. We train language-specific SentencePiece tokenizers on curated monolingual corpora, extend XLM-R's vocabulary with 30k Ge'ez-script subwords, and initialize their embeddings via subword averaging. VEXMLM undergoes two-stage training: (1) continued masked language modeling on the curated corpora and (2) supervised fine-tuning on question answering, named entity recognition, and sentiment analysis. VEXMLM substantially outperforms XLM-R and Glot500 across all evaluated tasks on Amharic and Tigrinya, with particularly strong gains on out-of-vocabulary entity recognition. Critically, improvements on Amharic and Tigrinya transfer to 17 languages in Africa. VEXMLM demonstrates that vocabulary expansion and tokenizer adaptation provide an effective, computationally efficient path to improve multilingual models for underrepresented languages without retraining from scratch. Resources: GitHub repository | Hugging Face models

cs.CL↗

arXiv2026

A Sovereign, Open-Source Foundation Model for German and English

We present Soofi S 30B-A3B, a sovereign, open-source Mixture-of-Experts (MoE) hybrid Mamba Transformer foundation model for German and English. Its hybrid design activates only 3B of 30B parameters per token and keeps the inference cache near-constant as context grows, giving it a decisive throughput advantage over dense models for long-context, high-concurrency deployment. Pretrained on roughly 27 trillion tokens with deliberately up-weighted German, Soofi S matches dense 14 to 27B models on aggregate English and German benchmarks while achieving the best code aggregates in both languages among 17 open base models, and outperforms every European sovereign baseline in our comparison, including ones far larger in active parameters. Among fully open models, Soofi S obtains the highest English and German evaluation scores, ahead of Olmo 3 32B and Apertus 70B. Soofi S was built end-to-end on the German Industrial AI Cloud, a sovereign HPC scale AI infrastructure operated by Deutsche Telekom in Munich. Soofi S will be released under highly permissive, open-access terms: weights, selected intermediate checkpoints, full per-source data accounting, hyperparameters, and training and evaluation code. Where source licenses permit, data-construction artifacts are released under permissive licenses; commercially licensed sources are documented with aggregate statistics and exact mixture accounting.

cs.CL↗

arXiv2026

Decomposing the Generalization Gap in PROTAC Activity Prediction: Variance Attribution and the Inter-Laboratory Ceiling

Machine-learning predictors of biochemical activity often exhibit large random-split-to-leave-one-target-out generalisation gaps that have been documented but not decomposed. We frame this as an evaluation-science question and use targeted protein degradation as the empirical test bed. PROTACs (proteolysis-targeting chimeras) are heterobifunctional small molecules that induce targeted protein degradation, with more than forty candidates currently in clinical trials; published predictors report AUROC of 0.85 to 0.91 under random-split cross-validation, while the leave-one-target-out (LOTO) protocol of Ribes et al. reduces performance to approximately 0.67. Random splits reward within-target interpolation, whereas LOTO measures the novel-target prediction that de-novo design depends on. We decompose this gap and identify inter-laboratory measurement variance as the dominant component, anchored by a within-target cross-laboratory cascade bounding the inter-laboratory contribution at 0.124 AUROC, well above the 0.05 contribution from binarisation-threshold choice. Across eight published architectures and ESM-2 protein language models up to 3B parameters, LOTO AUROC plateaus near 0.67, with a comparable plateau under SMILES-level deduplication; a 21-dimensional 2000-trial hyperparameter optimisation cannot break this ceiling, and the rank-1 single-seed configuration regresses by 0.161 AUROC under multi-seed validation, matching a closed-form selection-bias prediction (Bailey and Lopez de Prado, 2014). Few-shot k=5 stratified per-target retraining combined with ADMET features lifts 65-target LOTO AUROC from 0.668 to 0.7050, and post-hoc Platt scaling recovers raw output to within the 0.05 well-calibrated threshold. We release PROTAC-Bench (10,748 measurements, 173 targets, 65 LOTO folds), the variance-decomposition framework, the per-target calibration protocol, and the evaluation code.

cs.LG↗

arXiv2026

LLM Probe: Evaluating LLMs for Low-Resource Languages

Despite rapid advances in large language models (LLMs), their linguistic abilities in low-resource and morphologically rich languages are still not well understood due to limited annotated resources and the absence of standardized evaluation frameworks. This paper presents LLM Probe, a lexicon-based assessment framework designed to systematically evaluate the linguistic skills of LLMs in low-resource language environments. The framework analyzes models across four areas of language understanding: lexical alignment, part-of-speech recognition, morphosyntactic probing, and translation accuracy. To illustrate the framework, we create a manually annotated benchmark dataset using a low-resource Semitic language as a case study. The dataset comprises bilingual lexicons with linguistic annotations, including part-of-speech tags, grammatical gender, and morphosyntactic features, which demonstrate high inter-annotator agreement to ensure reliable annotations. We test a variety of models, including causal language models and sequence-to-sequence architectures. The results reveal notable differences in performance across various linguistic tasks: sequence-to-sequence models generally excel in morphosyntactic analysis and translation quality, whereas causal models demonstrate strong performance in lexical alignment but exhibit weaker translation accuracy. Our results emphasize the need for linguistically grounded evaluation to better understand LLM limitations in low-resource settings. We release LLM Probe and the accompanying benchmark dataset as open-source tools to promote reproducible benchmarking and to support the development of more inclusive multilingual language technologies.

cs.CL↗

arXiv2026

LGSE: Lexically Grounded Subword Embedding Initialization for Low-Resource Language Adaptation

Adapting pretrained language models to low-resource, morphologically rich languages remains a significant challenge. Existing vocabulary expansion methods typically rely on arbitrarily segmented subword units, resulting in fragmented lexical representations and loss of critical morphological information. To address this limitation, we propose the Lexically Grounded Subword Embedding Initialization (LGSE) framework, which introduces morphologically informed segmentation for initializing embeddings of novel tokens. Instead of using random vectors or arbitrary subwords, LGSE decomposes words into their constituent morphemes and constructs semantically coherent embeddings by averaging pretrained subword or FastText-based morpheme representations. When a token cannot be segmented into meaningful morphemes, its embedding is constructed using character n-gram representations to capture structural information. During Language-Adaptive Pretraining, we apply a regularization term that penalizes large deviations of newly introduced embeddings from their initialized values, preserving alignment with the original pretrained embedding space while enabling adaptation to the target language. To isolate the effect of initialization, we retain the original pre-trained model vocabulary and tokenizer and update only the new embeddings during adaptation. We evaluate LGSE on three NLP tasks: Question Answering, Named Entity Recognition, and Text Classification, in two morphologically rich, low-resource languages: Amharic and Tigrinya, where morphological segmentation resources are available. Experimental results show that LGSE consistently outperforms baseline methods across all tasks, demonstrating the effectiveness of morphologically grounded embedding initialization for improving representation quality in underrepresented languages. Project resources are available in the GitHub link.

cs.CL↗

arXiv2026

Cross-Modal Rationale Transfer for Explainable Humanitarian Classification on Social Media

Advances in social media data dissemination enable the provision of real-time information during a crisis. The information comes from different classes, such as infrastructure damages, persons missing or stranded in the affected zone, etc. Existing methods attempted to classify text and images into various humanitarian categories, but their decision-making process remains largely opaque, which affects their deployment in real-life applications. Recent work has sought to improve transparency by extracting textual rationales from tweets to explain predicted classes. However, such explainable classification methods have mostly focused on text, rather than crisis-related images. In this paper, we propose an interpretable-by-design multimodal classification framework. Our method first learns the joint representation of text and image using a visual language transformer model and extracts text rationales. Next, it extracts the image rationales via the mapping with text rationales. Our approach demonstrates how to learn rationales in one modality from another through cross-modal rationale transfer, which saves annotation effort. Finally, tweets are classified based on extracted rationales. Experiments are conducted over CrisisMMD benchmark dataset, and results show that our proposed method boosts the classification Macro-F1 by 2-35% while extracting accurate text tokens and image patches as rationales. Human evaluation also supports the claim that our proposed method is able to retrieve better image rationale patches (12%) that help to identify humanitarian classes. Our method adapts well to new, unseen datasets in zero-shot mode, achieving an accuracy of 80%.

cs.CL↗

arXiv2026

Training-Induced Bias Toward LLM-Generated Content in Dense Retrieval

Dense retrieval is a promising approach for acquiring relevant context or world knowledge in open-domain natural language processing tasks and is now widely used in information retrieval applications. However, recent reports claim a broad preference for text generated by large language models (LLMs). This bias is called "source bias", and it has been hypothesized that lower perplexity contributes to this effect. In this study, we revisit this claim by conducting a controlled evaluation to trace the emergence of such preferences across training stages and data sources. Using parallel human- and LLM-generated counterparts of the SciFact and Natural Questions (NQ320K) datasets, we compare unsupervised checkpoints with models fine-tuned using in-domain human text, in-domain LLM-generated text, and MS MARCO. Our results show the following: 1) Unsupervised retrievers do not exhibit a uniform pro-LLM preference. The direction and magnitude depend on the dataset. 2) Across the settings tested, supervised fine-tuning on MS MARCO consistently shifts the rankings toward LLM-generated text. 3) In-domain fine-tuning produces dataset-specific and inconsistent shifts in preference. 4) Fine-tuning on LLM-generated corpora induces a pronounced pro-LLM bias. Finally, a retriever-centric perplexity probe involving the reattachment of a language modeling head to the fine-tuned dense retriever encoder indicates agreement with relevance near chance, thereby weakening the explanatory power of perplexity. Our study demonstrates that source bias is a training-induced phenomenon rather than an inherent property of dense retrievers.

cs.IR↗

arXiv2026

Investigating the Impact of Histopathological Foundation Models on Regressive Prediction of Homologous Recombination Deficiency

Foundation models pretrained on large-scale histopathology data have found great success in various fields of computational pathology, but their impact on regressive biomarker prediction remains underexplored. In this work, we systematically evaluate histopathological foundation models for regression-based tasks, demonstrated through the prediction of homologous recombination deficiency (HRD) score - a critical biomarker for personalized cancer treatment. Within multiple instance learning frameworks, we extract patch-level features from whole slide images (WSI) using five state-of-the-art foundation models, and evaluate their impact compared to contrastive learning-based features. Models are trained to predict continuous HRD scores based on these extracted features across breast, endometrial, and lung cancer cohorts from two public medical data collections. Extensive experiments demonstrate that models trained on foundation model features consistently outperform the baseline in terms of predictive accuracy and generalization capabilities while exhibiting systematic differences among the foundation models. Additionally, we propose a distribution-based upsampling strategy to mitigate target imbalance in these datasets, significantly improving the recall and balanced accuracy for underrepresented but clinically important patient populations. Furthermore, we investigate the impact of different sampling strategies and instance bagsizes by ablation studies. Our results highlight the benefits of large-scale histopathological pretraining for more precise and transferable regressive biomarker prediction, showcasing its potential to advance AI-driven precision oncology.

cs.CV↗

arXiv2025

MedAI: Evaluating TxAgent's Therapeutic Agentic Reasoning in the NeurIPS CURE-Bench Competition

Therapeutic decision-making in clinical medicine constitutes a high-stakes domain in which AI guidance interacts with complex interactions among patient characteristics, disease processes, and pharmacological agents. Tasks such as drug recommendation, treatment planning, and adverse-effect prediction demand robust, multi-step reasoning grounded in reliable biomedical knowledge. Agentic AI methods, exemplified by TxAgent, address these challenges through iterative retrieval-augmented generation (RAG). TxAgent employs a fine-tuned Llama-3.1-8B model that dynamically generates and executes function calls to a unified biomedical tool suite (ToolUniverse), integrating FDA Drug API, OpenTargets, and Monarch resources to ensure access to current therapeutic information. In contrast to general-purpose RAG systems, medical applications impose stringent safety constraints, rendering the accuracy of both the reasoning trace and the sequence of tool invocations critical. These considerations motivate evaluation protocols treating token-level reasoning and tool-usage behaviors as explicit supervision signals. This work presents insights derived from our participation in the CURE-Bench NeurIPS 2025 Challenge, which benchmarks therapeutic-reasoning systems using metrics that assess correctness, tool utilization, and reasoning quality. We analyze how retrieval quality for function (tool) calls influences overall model performance and demonstrate performance gains achieved through improved tool-retrieval strategies. Our work was awarded the Excellence Award in Open Science. Complete information can be found at https://curebench.ai/.

cs.AI↗

arXiv2025

Towards Transparent Stance Detection: A Zero-Shot Approach Using Implicit and Explicit Interpretability

Zero-Shot Stance Detection (ZSSD) identifies the attitude of the post toward unseen targets. Existing research using contrastive, meta-learning, or data augmentation suffers from generalizability issues or lack of coherence between text and target. Recent works leveraging large language models (LLMs) for ZSSD focus either on improving unseen target-specific knowledge or generating explanations for stance analysis. However, most of these works are limited by their over-reliance on explicit reasoning, provide coarse explanations that lack nuance, and do not explicitly model the reasoning process, making it difficult to interpret the model's predictions. To address these issues, in our study, we develop a novel interpretable ZSSD framework, IRIS. We provide an interpretable understanding of the attitude of the input towards the target implicitly based on sequences within the text (implicit rationales) and explicitly based on linguistic measures (explicit rationales). IRIS considers stance detection as an information retrieval ranking task, understanding the relevance of implicit rationales for different stances to guide the model towards correct predictions without requiring the ground-truth of rationales, thus providing inherent interpretability. In addition, explicit rationales based on communicative features help decode the emotional and cognitive dimensions of stance, offering an interpretable understanding of the author's attitude towards the given target. Extensive experiments on the benchmark datasets of VAST, EZ-STANCE, P-Stance, and RFD using 50%, 30%, and even 10% training data prove the generalizability of our model, benefiting from the proposed architecture and interpretable design.

cs.CL↗

arXiv2025

Tokenization Disparities as Infrastructure Bias: How Subword Systems Create Inequities in LLM Access and Efficiency

Tokenization disparities pose a significant barrier to achieving equitable access to artificial intelligence across linguistically diverse populations. This study conducts a large-scale cross-linguistic evaluation of tokenization efficiency in over 200 languages to systematically quantify computational inequities in large language models (LLMs). Using a standardized experimental framework, we applied consistent preprocessing and normalization protocols, followed by uniform tokenization through the tiktoken library across all language samples. Comprehensive tokenization statistics were collected using established evaluation metrics, including Tokens Per Sentence (TPS) and Relative Tokenization Cost (RTC), benchmarked against English baselines. Our cross-linguistic analysis reveals substantial and systematic disparities: Latin-script languages consistently exhibit higher tokenization efficiency, while non-Latin and morphologically complex languages incur significantly greater token inflation, often 3-5 times higher RTC ratios. These inefficiencies translate into increased computational costs and reduced effective context utilization for underrepresented languages. Overall, the findings highlight structural inequities in current AI systems, where speakers of low-resource and non-Latin languages face disproportionate computational disadvantages. Future research should prioritize the development of linguistically informed tokenization strategies and adaptive vocabulary construction methods that incorporate typological diversity, ensuring more inclusive and computationally equitable multilingual AI systems.

cs.CL↗

arXiv2025

CliniBench: A Clinical Outcome Prediction Benchmark for Generative and Encoder-Based Language Models

With their growing capabilities, generative large language models (LLMs) are being increasingly investigated for complex medical tasks. However, their effectiveness in real-world clinical applications remains underexplored. To address this, we present CliniBench, the first benchmark that enables comparability of well-studied encoder-based classifiers and generative LLMs for discharge diagnosis prediction from admission notes in MIMIC-IV dataset. Our extensive study compares 12 generative LLMs and 3 encoder-based classifiers and demonstrates that encoder-based classifiers consistently outperform generative models in diagnosis prediction. We assess several retrieval augmentation strategies for in-context learning from similar patients and find that they provide notable performance improvements for generative LLMs.

cs.CL↗

arXiv2025

Low-Resource English-Tigrinya MT: Leveraging Multilingual Models, Custom Tokenizers, and Clean Evaluation Benchmarks

Despite advances in Neural Machine Translation (NMT), low-resource languages like Tigrinya remain underserved due to persistent challenges, including limited corpora, inadequate tokenization strategies, and the lack of standardized evaluation benchmarks. This paper investigates transfer learning techniques using multilingual pretrained models to enhance translation quality for morphologically rich, low-resource languages. We propose a refined approach that integrates language-specific tokenization, informed embedding initialization, and domain-adaptive fine-tuning. To enable rigorous assessment, we construct a high-quality, human-aligned English-Tigrinya evaluation dataset covering diverse domains. Experimental results demonstrate that transfer learning with a custom tokenizer substantially outperforms zero-shot baselines, with gains validated by BLEU, chrF, and qualitative human evaluation. Bonferroni correction is applied to ensure statistical significance across configurations. Error analysis reveals key limitations and informs targeted refinements. This study underscores the importance of linguistically aware modeling and reproducible benchmarks in bridging the performance gap for underrepresented languages. Resources are available at https://github.com/hailaykidu/MachineT_TigEng and https://huggingface.co/Hailay/MachineT_TigEng

cs.CL↗

arXiv2025

MoVoC: Morphology-Aware Subword Construction for Geez Script Languages

Subword-based tokenization methods often fail to preserve morphological boundaries, a limitation especially pronounced in low-resource, morphologically complex languages such as those written in the Geez script. To address this, we present MoVoC (Morpheme-aware Subword Vocabulary Construction) and train MoVoC-Tok, a tokenizer that integrates supervised morphological analysis into the subword vocabulary. This hybrid segmentation approach combines morpheme-based and Byte Pair Encoding (BPE) tokens to preserve morphological integrity while maintaining lexical meaning. To tackle resource scarcity, we curate and release manually annotated morpheme data for four Geez script languages and a morpheme-aware vocabulary for two of them. While the proposed tokenization method does not lead to significant gains in automatic translation quality, we observe consistent improvements in intrinsic metrics, MorphoScore, and Boundary Precision, highlighting the value of morphology-aware segmentation in enhancing linguistic fidelity and token efficiency. Our morpheme-annotated datasets and tokenizer will be publicly available to support further research in low-resource, morphologically rich languages. Our code and data are available on GitHub: https://github.com/hailaykidu/MoVoC

cs.CL↗

arXiv2025

"Where does it hurt?" -- Dataset and Study on Physician Intent Trajectories in Doctor Patient Dialogues

In a doctor-patient dialogue, the primary objective of physicians is to diagnose patients and propose a treatment plan. Medical doctors guide these conversations through targeted questioning to efficiently gather the information required to provide the best possible outcomes for patients. To the best of our knowledge, this is the first work that studies physician intent trajectories in doctor-patient dialogues. We use the `Ambient Clinical Intelligence Benchmark' (Aci-bench) dataset for our study. We collaborate with medical professionals to develop a fine-grained taxonomy of physician intents based on the SOAP framework (Subjective, Objective, Assessment, and Plan). We then conduct a large-scale annotation effort to label over 5000 doctor-patient turns with the help of a large number of medical experts recruited using Prolific, a popular crowd-sourcing platform. This large labeled dataset is an important resource contribution that we use for benchmarking the state-of-the-art generative and encoder models for medical intent classification tasks. Our findings show that our models understand the general structure of medical dialogues with high accuracy, but often fail to identify transitions between SOAP categories. We also report for the first time common trajectories in medical dialogue structures that provide valuable insights for designing `differential diagnosis' systems. Finally, we extensively study the impact of intent filtering for medical dialogue summarization and observe a significant boost in performance. We make the codes and data, including annotation guidelines, publicly available at https://github.com/DATEXIS/medical-intent-classification.

cs.CL↗

arXiv2025

Comply: Learning Sentences with Complex Weights inspired by Fruit Fly Olfaction

Biologically inspired neural networks offer alternative avenues to model data distributions. FlyVec is a recent example that draws inspiration from the fruit fly's olfactory circuit to tackle the task of learning word embeddings. Surprisingly, this model performs competitively even against deep learning approaches specifically designed to encode text, and it does so with the highest degree of computational efficiency. We pose the question of whether this performance can be improved further. For this, we introduce Comply. By incorporating positional information through complex weights, we enable a single-layer neural network to learn sequence representations. Our experiments show that Comply not only supersedes FlyVec but also performs on par with significantly larger state-of-the-art models. We achieve this without additional parameters. Comply yields sparse contextual representations of sentences that can be interpreted explicitly from the neuron weights.

cs.CL↗

arXiv2025

Enhancing Offline Reinforcement Learning with Curriculum Learning-Based Trajectory Valuation

The success of deep reinforcement learning (DRL) relies on the availability and quality of training data, often requiring extensive interactions with specific environments. In many real-world scenarios, where data collection is costly and risky, offline reinforcement learning (RL) offers a solution by utilizing data collected by domain experts and searching for a batch-constrained optimal policy. This approach is further augmented by incorporating external data sources, expanding the range and diversity of data collection possibilities. However, existing offline RL methods often struggle with challenges posed by non-matching data from these external sources. In this work, we specifically address the problem of source-target domain mismatch in scenarios involving mixed datasets, characterized by a predominance of source data generated from random or suboptimal policies and a limited amount of target data generated from higher-quality policies. To tackle this problem, we introduce Transition Scoring (TS), a novel method that assigns scores to transitions based on their similarity to the target domain, and propose Curriculum Learning-Based Trajectory Valuation (CLTV), which effectively leverages these transition scores to identify and prioritize high-quality trajectories through a curriculum learning approach. Our extensive experiments across various offline RL methods and MuJoCo environments, complemented by rigorous theoretical analysis, demonstrate that CLTV enhances the overall performance and transferability of policies learned by offline RL algorithms.

cs.LG↗

arXiv2024

FlexTSF: A Flexible Forecasting Model for Time Series with Variable Regularities

Forecasting time series with irregular temporal structures remains challenging for universal pre-trained models. Existing approaches often assume regular sampling or depend heavily on imputation, limiting their applicability in real-world scenarios where irregularities are prevalent due to diverse sensing devices and recording practices. We introduce FlexTSF, a flexible forecasting model specifically designed for time series data with variable temporal regularities. At its foundation lies the IVP Patcher, a continuous-time patching module leveraging Initial Value Problems (IVPs) to inherently support uneven time intervals, variable sequence lengths, and missing values. FlexTSF employs a decoder-only architecture that integrates normalized timestamp inputs and domain-specific statistics through a specialized causal self-attention mechanism, enabling adaptability across domains. Extensive experiments on 16 datasets demonstrate FlexTSF's effectiveness, significantly outperforming existing models in classic forecasting scenarios, zero-shot generalization, and low-resource fine-tuning conditions. Ablation studies confirm the contributions of each design component and the advantage of not relying on predefined fixed patch lengths.

cs.LG↗