Update README.md

README.md

@@ -30,20 +30,20 @@ The native component combines curated datasets and web-scraped French medical re
 
 - **[S-Editions](https://s-editions.fr/)** – 526 question–answer pairs from a French educational platform for medical students.  
 
-- **[MediQAl](https://arxiv.org/abs/2501.01234)** ~\cite{bazoge2025mediqal}:  
+- **[MediQAl](https://github.com/abazoge/MediQAl)**:  
   32 603 items from national medical examinations covering 41 medical specialties.  
 
-- **[FrenchMedMCQA](https://huggingface.co/datasets/ylabrak/FrenchMedMCQA)** ~\cite{labrak2023frenchmedmcqa}:  
+- **[FrenchMedMCQA](https://arxiv.org/abs/2304.04280)**:  
   3 105 pharmacy-focused multiple-choice questions.  
 
 - **[mlabonne/medical-cases-fr](https://huggingface.co/datasets/mlabonne/medical-cases-fr)** and **[mlabonne/medical-mcqa-fr](https://huggingface.co/datasets/mlabonne/medical-mcqa-fr)** ~\cite{mlabonne_medical_cases_fr, mlabonne_medical_mqca_fr}:  
   12 194 examples originating from French medical exam databases.  
 
-- **[FrBMedQA](https://huggingface.co/datasets/LIUM-NLP/FrBMedQA)** ~\cite{kaddari2022frbmedqa}:  
+- **[FrBMedQA](https://www.researchgate.net/publication/365908538_FrBMedQA_the_first_French_biomedical_question_answering_dataset)**:  
   19 836 questions derived from French biomedical Wikipedia articles spanning eight UMLS semantic groups (chemicals and drugs, anatomy, physiology, disorders, phenomena, procedures, genes and molecular sequences, and devices).  
-  Originally closed-form, these questions were reformulated into multiple-choice format using *[GPT-4o-mini](https://openai.com/research/gpt-4o)* ~\cite{hurst2024gpt} for standardization.  
+  Originally closed-form, these questions were reformulated into multiple-choice format using *[GPT-4o-mini](https://openai.com/research/gpt-4o)* for standardization.  
 
-- **[Biomedical Translation Corpora (WMT)](https://www.statmt.org/wmt23/biomedical-task.html)** ~\cite{biomedical_translation_corpora}:  
+- **[Biomedical Translation Corpora (WMT)](https://github.com/biomedical-translation-corpora/corpora)**:  
   Bilingual biomedical translation data from the WMT challenge repositories were reformulated into instruction–response pairs.  
   Each instruction requests the **French translation** of an English biomedical passage, reframing translation as an instruction-following task aligned with the native portion.
 
@@ -55,15 +55,6 @@ Used to fine-tune biomedical LLMs for:
 - Evaluating cross-domain instruction generalization
 
 
-## References
-
-- Bazoge, A. et al. (2025). *MediQAl: A Large-Scale French Biomedical Question Answering Dataset*. [arXiv:2501.01234](https://arxiv.org/abs/2501.01234)  
-- Labrak, Y. et al. (2023). *FrenchMedMCQA: A French Multiple-Choice Question Answering Dataset for Medical Education*. [Hugging Face](https://huggingface.co/datasets/ylabrak/FrenchMedMCQA)  
-- Kaddari, S. et al. (2022). *FrBMedQA: A French Biomedical Question Answering Benchmark*. [Hugging Face](https://huggingface.co/datasets/LIUM-NLP/FrBMedQA)  
-- Hurst, N. et al. (2024). *GPT-4o-mini: Efficient Instruction-Tuned Multimodal Foundation Model*. [OpenAI Research](https://openai.com/research/gpt-4o)  
-- Biomedical Translation Corpora (WMT Challenge Data, 2016–2024). [WMT Biomedical Task](https://www.statmt.org/wmt23/biomedical-task.html)  
-- Mlabonne, T. (2023). *medical-cases-fr* and *medical-mcqa-fr*. [Hugging Face](https://huggingface.co/mlabonne)  
-- S-Editions (2023). *French Medical Exam Practice Platform*. [Website](https://s-editions.fr/)