Papers by Mihai Dascalu
How Hard is this Test Set? NLI Characterization by Exploiting Training Dynamics (2024.emnlp-main)
Copied to clipboard
| Challenge: | Popular datasets suffer from systematic spurious correlations that artificially inflate actual model performance. |
| Approach: | They propose a method for the automated creation of a challenging test set without relying on manual construction of artificial and unrealistic examples. |
| Outcome: | The proposed method reduces spurious correlations and improves model performance . examples labeled as having the highest difficulty show markedly decreased performance compared to the full dataset . |
RoBERT – A Romanian BERT Model (2020.coling-main)
Copied to clipboard
| Challenge: | Existing pre-trained language models learn contextualized representations by using unlabeled text data and obtain state of the art results on a multitude of NLP tasks. |
| Approach: | They propose a pre-trained BERT model for Romanian language processing and compare it with multi-lingual models on seven Romanian specific NLP tasks. |
| Outcome: | The proposed model outperforms multi-lingual models on seven Romanian specific NLP tasks on sentiment analysis, dialect and cross-dialect topic identification, and diacritics restoration. |
“Vorbești Românește?” A Recipe to Train Powerful Romanian LLMs with English Instructions (2024.findings-emnlp)
Copied to clipboard
Mihai Masala, Denis Ilie-Ablachim, Alexandru Dima, Dragos Georgian Corlatescu, Miruna-Andreea Zavelca, Ovio Olaru, Simina-Maria Terian, Andrei Terian, Marius Leordeanu, Horia Velicu, Marius Popescu, Mihai Dascalu, Traian Rebedea
| Challenge: | Large Language Models (LLMs) have achieved almost human-like performance on various tasks. |
| Approach: | They 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. |
| Outcome: | The proposed model trains, evaluates and releases open-source models tailored for Romanian. |
ORSO QGen: Odds-Ratio Steerable Optimization for Controlling Question Generation (2026.findings-eacl)
Copied to clipboard
| Challenge: | Existing methods for question generation rely on supervised fine-tuning . scarcity of datasets with multiple fine-grained attributes limits ability of models to generalize or transfer control to new combinations of attributes. |
| Approach: | They propose a framework to enhance attribute sensitivity in question generation models . they propose supervised fine-tuning with explicit attribute labels to produce questions that fit predefined characteristics. |
| Outcome: | The proposed framework improves attribute sensitivity while maintaining quality of output. |
Domain Adaptation in Multilingual and Multi-Domain Monolingual Settings for Complex Word Identification (2022.acl-long)
Copied to clipboard
| Challenge: | Existing datasets for complex word identification (CWI) are limited and the difficulty of the task is augmented by the scarcity of input examples. |
| Approach: | They propose a novel training technique for the complex word identification task based on domain adaptation to improve character and context representations. |
| Outcome: | The proposed training technique improves the target character and context representations and also smooths differences between datasets. |
Distilling the Knowledge of Romanian BERTs Using Multiple Teachers (2022.lrec-1)
Copied to clipboard
Andrei-Marius Avram, Darius Catrina, Dumitru-Clementin Cercel, Mihai Dascalu, Traian Rebedea, Vasile Pais, Dan Tufis
| Challenge: | Existing approaches to train pre-trained language models focus on the English language, thus widening the gap when considering low-resource languages. |
| Approach: | They propose three versions of distilled BERT models for the Romanian language . they argue that the models offer performance comparable to their teachers . |
| Outcome: | The proposed models perform comparable to their teachers, while being twice as fast on a GPU and 35% smaller. |
Towards Building the LEMI Readability Platform for Children’s Literature in the Romanian Language (2024.lrec-main)
Copied to clipboard
Madalina Chitez, Mihai Dascalu, Aura Cristina Udrea, Cosmin Strilețchi, Karla Csürös, Roxana Rogobete, Alexandru Oravițan
| Challenge: | Currently, no existing platform integrates a research-based readability formula for the Romanian language, making this tool unique. |
| Approach: | They propose a new readability tool for children’s literature in the Romanian language that uses a self-compiled corpus and a text analysis interface to generate automatic readability reports for uploaded short texts. |
| Outcome: | The proposed readability tool is specifically targeted at primary school students aged 7-11 . it extracts, tests, and calibrates a readability formula for Romanian using the children’s literature corpus and the platform functionalities. |
The Strawberry Problem: Emergence of Character-level Understanding in Tokenized Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large Language Models fail at simple character-level tasks due to low mutual information, study finds . authors propose a lightweight architectural modification that improves character- level reasoning . |
| Approach: | They propose a lightweight architectural modification that improves character-level reasoning while preserving the inductive advantages of subword models. |
| Outcome: | The proposed model improves character-level reasoning while preserving the advantages of subword models. |