Language models (LMs) have demonstrated remarkable capabilities in NLP, yet adapting them efficiently and robustly to specific tasks remains challenging. As their scale and complexity grow, fine-tuning LMs on labelled data often underutilizes available unlabelled data, leads to overfitting on small task-specific sets, and imposes significant computational costs. These limitations hamper their application to the open-ended landscape of real-world language tasks. This thesis proposes a series of methods to better adapt LMs to downstream applications. First, we explore strategies for extracting task-relevant knowledge from unlabelled data, introducing a novel continued pre-training technique that outperforms state-of-the-art semi-supervised approaches. Next, we present a parameter-efficient fine-tuning method that substantially reduces memory and compute costs while maintaining competitive performance. We also introduce improved supervised fine-tuning methods that enable LMs to better follow instructions, especially when labelled data is scarce, enhancing their performance across a range of NLP tasks, including open-ended generation. Finally, we develop new evaluation methods and benchmarks, such as multi-hop spatial reasoning tasks, to assess LM capabilities and adaptation more comprehensively. Through extensive empirical studies across diverse NLP tasks, our results demonstrate that these approaches substantially improve LM robustness, efficiency, and generalization, making them more adaptable to a broad range of applications. These advances mark a significant step towards more robust and efficient LMs, bringing us closer to the goal of artificial general intelligence.

Multi-lingual competence in large language models is often evaluated via static data benchmarks such as Belebele, M-MMLU and M-GSM. However, these evaluations often fail to provide an adequate understanding of the practical performance and robustness of models across multi-lingual settings. In response, we create multi-lingual functional benchmarks -- Cross-Lingual Grade School Math Symbolic (CL-GSM Symbolic) and Cross-Lingual Instruction-Following Eval (CL-IFEval)-- by translating existing functional benchmark templates from English to five additional languages that span the range of resources available for NLP: French, Spanish, Hindi, Arabic and Yoruba. Our results reveal that some static multi-lingual benchmarks capture functional performance much more closely than others (i.e. across models, there is a 24%, 17% and 18% decrease in performance between M-GSM and CL-GSM Symbolic in English, French and Spanish respectively; similarly there's a 15 - 24% performance drop across languages between Belebele and CL-IFEval, and only a 0.5% to 3% performance drop between M-MMLU and CL-IFEval). Similarly, we find that model robustness across languages varies significantly, with certain languages (eg. Arabic, English) being the most consistently well performing across evaluation iterations.

In this work, we introduce skLEP, the first comprehensive benchmark specifically designed for evaluating Slovak natural language understanding (NLU) models. We have compiled skLEP to encompass nine diverse tasks that span token-level, sentence-pair, and document-level challenges, thereby offering a thorough assessment of model capabilities. To create this benchmark, we curated new, original datasets tailored for Slovak and meticulously translated established English NLU resources. Within this paper, we also present the first systematic and extensive evaluation of a wide array of Slovak-specific, multilingual, and English pre-trained language models using the skLEP tasks. Finally, we also release the complete benchmark data, an open-source toolkit facilitating both fine-tuning and evaluation of models, and a public leaderboard at https://github.com/slovak-nlp/sklep in the hopes of fostering reproducibility and drive future research in Slovak NLU.

In this paper, we compare Czech-specific and multilingual sentence embedding models through intrinsic and extrinsic evaluation paradigms. For intrinsic evaluation, we employ Costra, a complex sentence transformation dataset, and several Semantic Textual Similarity (STS) benchmarks to assess the ability of the embeddings to capture linguistic phenomena such as semantic similarity, temporal aspects, and stylistic variations. In the extrinsic evaluation, we fine-tune each embedding model using COMET-based metrics for machine translation evaluation. Our experiments reveal an interesting disconnect: models that excel in intrinsic semantic similarity tests do not consistently yield superior performance on downstream translation evaluation tasks. Conversely, models with seemingly over-smoothed embedding spaces can, through fine-tuning, achieve excellent results. These findings highlight the complex relationship between semantic property probes and downstream task, emphasizing the need for more research into 'operationalizable semantics' in sentence embeddings, or more in-depth downstream tasks datasets (here translation evaluation)

In Machine Translation (MT) evaluation, metric performance is assessed based on agreement with human judgments. In recent years, automatic metrics have demonstrated increasingly high levels of agreement with humans. To gain a clearer understanding of metric performance and establish an upper bound, we incorporate human baselines in the MT meta-evaluation, that is, the assessment of MT metrics' capabilities. Our results show that human annotators are not consistently superior to automatic metrics, with state-of-the-art metrics often ranking on par with or higher than human baselines. Despite these findings suggesting human parity, we discuss several reasons for caution. Finally, we explore the broader implications of our results for the research field, asking: Can we still reliably measure improvements in MT evaluation? With this work, we aim to shed light on the limits of our ability to measure progress in the field, fostering discussion on an issue that we believe is crucial to the entire MT evaluation community.

Lemmatization is crucial for NLP tasks in morphologically rich languages with ambiguous orthography like Arabic, but existing tools face challenges due to inconsistent standards and limited genre coverage. This paper introduces two novel approaches that frame lemmatization as classification into a Lemma-POS-Gloss (LPG) tagset, leveraging machine translation and semantic clustering. We also present a new Arabic lemmatization test set covering diverse genres, standardized alongside existing datasets. We evaluate character level sequence-to-sequence models, which perform competitively and offer complementary value, but are limited to lemma prediction (not LPG) and prone to hallucinating implausible forms. Our results show that classification and clustering yield more robust, interpretable outputs, setting new benchmarks for Arabic lemmatization.

This work describes the participation of the MLLP-VRAIN research group in the shared task of the IWSLT 2025 Simultaneous Speech Translation track. Our submission addresses the unique challenges of real-time translation of long-form speech by developing a modular cascade system that adapts strong pre-trained models to streaming scenarios. We combine Whisper Large-V3-Turbo for ASR with the multilingual NLLB-3.3B model for MT, implementing lightweight adaptation techniques rather than training new end-to-end models from scratch. Our approach employs document-level adaptation with prefix training to enhance the MT model's ability to handle incomplete inputs, while incorporating adaptive emission policies including a wait-$k$ strategy and RALCP for managing the translation stream. Specialized buffer management techniques and segmentation strategies ensure coherent translations across long audio sequences. Experimental results on the ACL60/60 dataset demonstrate that our system achieves a favorable balance between translation quality and latency, with a BLEU score of 31.96 and non-computational-aware StreamLAAL latency of 2.94 seconds. Our final model achieves a preliminary score on the official test set (IWSLT25Instruct) of 29.8 BLEU. Our work demonstrates that carefully adapted pre-trained components can create effective simultaneous translation systems for long-form content without requiring extensive in-domain parallel data or specialized end-to-end training.

Machine translation (MT) post-editing and research data collection often rely on inefficient, disconnected workflows. We introduce TranslationCorrect, an integrated framework designed to streamline these tasks. TranslationCorrect combines MT generation using models like NLLB, automated error prediction using models like XCOMET or LLM APIs (providing detailed reasoning), and an intuitive post-editing interface within a single environment. Built with human-computer interaction (HCI) principles in mind to minimize cognitive load, as confirmed by a user study. For translators, it enables them to correct errors and batch translate efficiently. For researchers, TranslationCorrect exports high-quality span-based annotations in the Error Span Annotation (ESA) format, using an error taxonomy inspired by Multidimensional Quality Metrics (MQM). These outputs are compatible with state-of-the-art error detection models and suitable for training MT or post-editing systems. Our user study confirms that TranslationCorrect significantly improves translation efficiency and user satisfaction over traditional annotation methods.

Entity-aware machine translation (EAMT) is a complicated task in natural language processing due to not only the shortage of translation data related to the entities needed to translate but also the complexity in the context needed to process while translating those entities. In this paper, we propose a method that applies multi-task learning to optimize the performance of the two subtasks named entity recognition and machine translation, which improves the final performance of the Entity-aware machine translation task. The result and analysis are performed on the dataset provided by the organizer of Task 2 of the SemEval 2025 competition.

Past work has long recognized the important role of context in guiding how humans search their memory. While context-based memory models can explain many memory phenomena, it remains unclear why humans develop such architectures over possible alternatives in the first place. In this work, we demonstrate that foundational architectures in neural machine translation -- specifically, recurrent neural network (RNN)-based sequence-to-sequence models with attention -- exhibit mechanisms that directly correspond to those specified in the Context Maintenance and Retrieval (CMR) model of human memory. Since neural machine translation models have evolved to optimize task performance, their convergence with human memory models provides a deeper understanding of the functional role of context in human memory, as well as presenting new ways to model human memory. Leveraging this convergence, we implement a neural machine translation model as a cognitive model of human memory search that is both interpretable and capable of capturing complex dynamics of learning. We show that our model accounts for both averaged and optimal human behavioral patterns as effectively as context-based memory models. Further, we demonstrate additional strengths of the proposed model by evaluating how memory search performance emerges from the interaction of different model components.