Mihai Dascalu

Mihai Masala, Marius Leordeanu, Mihai Dascalu et al.

Vision-Language Models (VLMs) largely follow the text-only LLM trajectory, excelling on English benchmarks but sharply degrading on low-resource languages, where neither large-scale image-text corpora nor culturally grounded evaluations exist. We present a systematic study of building a language-specific VLM for Romanian, covering the full pipeline from data construction to architectural choices. We translate established English VLM training and evaluation corpora into Romanian, applying machine translation to textual annotations and to in-image text, preserving visual grounding while adapting the textual content. Using this data, we train and ablate a series of VLMs to isolate the contribution of (i) vision backbones of varying scale and pretraining, (ii) language backbones from multilingual to Romanian-adapted LLMs, and (iii) OCR-style image-text data. We further curate HoraVQA, a culturally native evaluation set grounded in Romanian everyday scenes. Romanian-adapted VLMs consistently outperform their same-sized counterparts and, across all evaluated benchmarks, even surpass models from the next larger size category.

Andrei Paraschiv, Mihai Dascalu, Dumitru-Clementin Cercel

In recent times, the detection of hate-speech, offensive, or abusive language in online media has become an important topic in NLP research due to the exponential growth of social media and the propagation of such messages, as well as their impact. Misogyny detection, even though it plays an important part in hate-speech detection, has not received the same attention. In this paper, we describe our classification systems submitted to the SemEval-2022 Task 5: MAMI - Multimedia Automatic Misogyny Identification. The shared task aimed to identify misogynous content in a multi-modal setting by analysing meme images together with their textual captions. To this end, we propose two models based on the pre-trained UNITER model, one enhanced with an image sentiment classifier, whereas the second leverages a Vocabulary Graph Convolutional Network (VGCN). Additionally, we explore an ensemble using the aforementioned models. Our best model reaches an F1-score of 71.4% in Sub-task A and 67.3% for Sub-task B positioning our team in the upper third of the leaderboard. We release the code and experiments for our models on GitHub

Hunter McNichols, Alexander Scarlatos, Mihai Dascalu et al.

An effective method of teaching across disciplines is to provide examples of high-quality work. However, an example may be significantly different from a student's current work, making it challenging for them to emulate. An ideal learning demonstration is a counterfactual version of the student work, an improved version that is still similar to their own. Existing automated approaches for counterfactual text generation using Large Language Models (LLMs) result in domain-specific systems that are difficult to translate into practical applications. We present the Gumbel Machine, a flexible, modular approach to generating counterfactuals that leverages LLM instruction-following capabilities while encouraging similarity to a reference factual text. Central to our approach is a novel, controlled decoding algorithm, $β$-Hindsight control, which uses latent randomness as a tunable similarity control mechanism during counterfactual generation. Experiments on datasets of student writing, scored on various criteria, demonstrate the effectiveness of our approach at generating counterfactuals both rubric-consistent and similar to a reference.

George-Eduard Zaharia, Răzvan-Alexandru Smădu, Dumitru-Clementin Cercel et al.

Complex word identification (CWI) is a cornerstone process towards proper text simplification. CWI is highly dependent on context, whereas its difficulty is augmented by the scarcity of available datasets which vary greatly in terms of domains and languages. As such, it becomes increasingly more difficult to develop a robust model that generalizes across a wide array of input examples. In this paper, we propose a novel training technique for the CWI task based on domain adaptation to improve the target character and context representations. This technique addresses the problem of working with multiple domains, inasmuch as it creates a way of smoothing the differences between the explored datasets. Moreover, we also propose a similar auxiliary task, namely text simplification, that can be used to complement lexical complexity prediction. Our model obtains a boost of up to 2.42% in terms of Pearson Correlation Coefficients in contrast to vanilla training techniques, when considering the CompLex from the Lexical Complexity Prediction 2021 dataset. At the same time, we obtain an increase of 3% in Pearson scores, while considering a cross-lingual setup relying on the Complex Word Identification 2018 dataset. In addition, our model yields state-of-the-art results in terms of Mean Absolute Error.

Răzvan-Alexandru Smădu, George-Eduard Zaharia, Andrei-Marius Avram et al.

Keyphrase identification and classification is a Natural Language Processing and Information Retrieval task that involves extracting relevant groups of words from a given text related to the main topic. In this work, we focus on extracting keyphrases from scientific documents. We introduce TA-DA, a Topic-Aware Domain Adaptation framework for keyphrase extraction that integrates Multi-Task Learning with Adversarial Training and Domain Adaptation. Our approach improves performance over baseline models by up to 5% in the exact match of the F1-score.

Andrei Paraschiv, Mihai Dascalu

Conspiracy theories have become a prominent and concerning aspect of online discourse, posing challenges to information integrity and societal trust. As such, we address conspiracy theory detection as proposed by the ACTI @ EVALITA 2023 shared task. The combination of pre-trained sentence Transformer models and data augmentation techniques enabled us to secure first place in the final leaderboard of both sub-tasks. Our methodology attained F1 scores of 85.71% in the binary classification and 91.23% for the fine-grained conspiracy topic classification, surpassing other competing systems.

Andreea Dutulescu, Stefan Ruseti, Mihai Dascalu

Transformer-based language models often achieve strong results on mathematical reasoning benchmarks while remaining fragile on basic numerical understanding and arithmetic operations. A central limitation is that numbers are processed as symbolic tokens whose embeddings do not explicitly encode numerical value, leading to systematic errors. We introduce a value-aware numerical representation that augments standard tokenized inputs with a dedicated prefix token whose embedding is explicitly conditioned on the underlying numerical value. This mechanism injects magnitude information directly into the model's input space while remaining compatible with existing tokenizers and decoder-only Transformer architectures. Evaluation on arithmetic tasks shows that the proposed approach outperforms baselines across numerical formats, tasks, and operand lengths. These results indicate that explicitly encoding numerical value is an effective and efficient way to improve fundamental numerical robustness in language models.

Adrian Cosma, Stefan Ruseti, Emilian Radoi et al.

Despite their remarkable progress across diverse domains, Large Language Models (LLMs) consistently fail at simple character-level tasks, such as counting letters in words, due to a fundamental limitation: tokenization. In this work, we frame this limitation as a problem of low mutual information and analyze it in terms of concept emergence. Using a suite of 19 synthetic tasks that isolate character-level reasoning in a controlled setting, we show that such capabilities emerge suddenly and only late in training. We find that percolation-based models of concept emergence explain these patterns, suggesting that learning character composition is not fundamentally different from learning commonsense knowledge. To address this bottleneck, we propose a lightweight architectural modification that significantly improves character-level reasoning while preserving the inductive advantages of subword models. Together, our results bridge low-level perceptual gaps in tokenized LMs and provide a principled framework for understanding and mitigating their structural blind spots. We make our code publicly available.

Mihai Masala, Denis C. Ilie-Ablachim, Alexandru Dima et al.

In recent years, Large Language Models (LLMs) have achieved almost human-like performance on various tasks. While some LLMs have been trained on multilingual data, most of the training data is in English; hence, their performance in English greatly exceeds other languages. To our knowledge, we are the first to collect and translate a large collection of texts, instructions, and benchmarks and train, evaluate, and release open-source LLMs tailored for Romanian. We evaluate our methods on four different categories, including academic benchmarks, MT-Bench (manually translated), and a professionally built historical, cultural, and social benchmark adapted to Romanian. We argue for the usefulness and high performance of RoLLMs by obtaining state-of-the-art results across the board. We publicly release all resources (i.e., data, training and evaluation code, models) to support and encourage research on Romanian LLMs while concurrently creating a generalizable recipe, adequate for other low or less-resourced languages.

Teodor-Mihai Cotet, Stefan Ruseti, Mihai Dascalu

Resources for Grammatical Error Correction (GEC) in non-English languages are scarce, while available spellcheckers in these languages are mostly limited to simple corrections and rules. In this paper we introduce a first GEC corpus for Romanian consisting of 10k pairs of sentences. In addition, the German version of ERRANT (ERRor ANnotation Toolkit) scorer was adapted for Romanian to analyze this corpus and extract edits needed for evaluation. Multiple neural models were experimented, together with pretraining strategies, which proved effective for GEC in low-resource settings. Our baseline consists of a small Transformer model trained only on the GEC dataset (F0.5 of 44.38), whereas the best performing model is produced by pretraining a larger Transformer model on artificially generated data, followed by finetuning on the actual corpus (F0.5 of 53.76). The proposed method for generating additional training examples is easily extensible and can be applied to any language, as it requires only a POS tagger

Mihai Masala, Denis C. Ilie-Ablachim, Dragos Corlatescu et al.

In recent years, Large Language Models (LLMs) have achieved almost human-like performance on various tasks. While some LLMs have been trained on multilingual data, most of the training data is in English. Hence, their performance in English greatly exceeds their performance in other languages. This document presents our approach to training and evaluating the first foundational and chat LLM specialized for Romanian.

Andrei Paraschiv, Dumitru-Clementin Cercel, Mihai Dascalu

The real-world impact of polarization and toxicity in the online sphere marked the end of 2020 and the beginning of this year in a negative way. Semeval-2021, Task 5 - Toxic Spans Detection is based on a novel annotation of a subset of the Jigsaw Unintended Bias dataset and is the first language toxicity detection task dedicated to identifying the toxicity-level spans. For this task, participants had to automatically detect character spans in short comments that render the message as toxic. Our model considers applying Virtual Adversarial Training in a semi-supervised setting during the fine-tuning process of several Transformer-based models (i.e., BERT and RoBERTa), in combination with Conditional Random Fields. Our approach leads to performance improvements and more robust models, enabling us to achieve an F1-score of 65.73% in the official submission and an F1-score of 66.13% after further tuning during post-evaluation.

George-Eduard Zaharia, Dumitru-Clementin Cercel, Mihai Dascalu

Reading is a complex process which requires proper understanding of texts in order to create coherent mental representations. However, comprehension problems may arise due to hard-to-understand sections, which can prove troublesome for readers, while accounting for their specific language skills. As such, steps towards simplifying these sections can be performed, by accurately identifying and evaluating difficult structures. In this paper, we describe our approach for the SemEval-2021 Task 1: Lexical Complexity Prediction competition that consists of a mixture of advanced NLP techniques, namely Transformer-based language models, pre-trained word embeddings, Graph Convolutional Networks, Capsule Networks, as well as a series of hand-crafted textual complexity features. Our models are applicable on both subtasks and achieve good performance results, with a MAE below 0.07 and a Person correlation of .73 for single word identification, as well as a MAE below 0.08 and a Person correlation of .79 for multiple word targets. Our results are just 5.46% and 6.5% lower than the top scores obtained in the competition on the first and the second subtasks, respectively.

Răzvan-Alexandru Smădu, Dumitru-Clementin Cercel, Mihai Dascalu

Detecting humor is a challenging task since words might share multiple valences and, depending on the context, the same words can be even used in offensive expressions. Neural network architectures based on Transformer obtain state-of-the-art results on several Natural Language Processing tasks, especially text classification. Adversarial learning, combined with other techniques such as multi-task learning, aids neural models learn the intrinsic properties of data. In this work, we describe our adversarial multi-task network, AMTL-Humor, used to detect and rate humor and offensive texts from Task 7 at SemEval-2021. Each branch from the model is focused on solving a related task, and consists of a BiLSTM layer followed by Capsule layers, on top of BERTweet used for generating contextualized embeddings. Our best model consists of an ensemble of all tested configurations, and achieves a 95.66% F1-score and 94.70% accuracy for Task 1a, while obtaining RMSE scores of 0.6200 and 0.5318 for Tasks 1b and 2, respectively.

Andrei-Marius Avram, George-Eduard Zaharia, Dumitru-Clementin Cercel et al.

Extracting semantic information on measurements and counts is an important topic in terms of analyzing scientific discourses. The 8th task of SemEval-2021: Counts and Measurements (MeasEval) aimed to boost research in this direction by providing a new dataset on which participants train their models to extract meaningful information on measurements from scientific texts. The competition is composed of five subtasks that build on top of each other: (1) quantity span identification, (2) unit extraction from the identified quantities and their value modifier classification, (3) span identification for measured entities and measured properties, (4) qualifier span identification, and (5) relation extraction between the identified quantities, measured entities, measured properties, and qualifiers. We approached these challenges by first identifying the quantities, extracting their units of measurement, classifying them with corresponding modifiers, and afterwards using them to jointly solve the last three subtasks in a multi-turn question answering manner. Our best performing model obtained an overlapping F1-score of 36.91% on the test set.

George-Eduard Zaharia, Dumitru-Clementin Cercel, Mihai Dascalu

Complex Word Identification (CWI) is a task centered on detecting hard-to-understand words, or groups of words, in texts from different areas of expertise. The purpose of CWI is to highlight problematic structures that non-native speakers would usually find difficult to understand. Our approach uses zero-shot, one-shot, and few-shot learning techniques, alongside state-of-the-art solutions for Natural Language Processing (NLP) tasks (i.e., Transformers). Our aim is to provide evidence that the proposed models can learn the characteristics of complex words in a multilingual environment by relying on the CWI shared task 2018 dataset available for four different languages (i.e., English, German, Spanish, and also French). Our approach surpasses state-of-the-art cross-lingual results in terms of macro F1-score on English (0.774), German (0.782), and Spanish (0.734) languages, for the zero-shot learning scenario. At the same time, our model also outperforms the state-of-the-art monolingual result for German (0.795 macro F1-score).

Andrei Paraschiv, Dumitru-Clementin Cercel, Mihai Dascalu

Manipulative and misleading news have become a commodity for some online news outlets and these news have gained a significant impact on the global mindset of people. Propaganda is a frequently employed manipulation method having as goal to influence readers by spreading ideas meant to distort or manipulate their opinions. This paper describes our participation in the SemEval-2020, Task 11: Detection of Propaganda Techniques in News Articles competition. Our approach considers specializing a pre-trained BERT model on propagandistic and hyperpartisan news articles, enabling it to create more adequate representations for the two subtasks, namely propaganda Span Identification (SI) and propaganda Technique Classification (TC). Our proposed system achieved a F1-score of 46.060% in subtask SI, ranking 5th in the leaderboard from 36 teams and a micro-averaged F1 score of 54.302% for subtask TC, ranking 19th from 32 teams.

Stefan Ruseti, Teodor-Mihai Cotet, Mihai Dascalu

Diacritics restoration is a mandatory step for adequately processing Romanian texts, and not a trivial one, as you generally need context in order to properly restore a character. Most previous methods which were experimented for Romanian restoration of diacritics do not use neural networks. Among those that do, there are no solutions specifically optimized for this particular language (i.e., they were generally designed to work on many different languages). Therefore we propose a novel neural architecture based on recurrent neural networks that can attend information at different levels of abstractions in order to restore diacritics.