L’alignement des modèles de langage avec les valeurs humaines est essentiel, à mesure qu’ils s’intègrent dans la vie quotidienne. Ces modèles sont souvent adaptés aux préférences des utilisateurs mais il est important de veiller à ce qu’ils respectent des normes morales en situation réelle. Malgré des avancées dans d’autres langues, le raisonnement moral des modèles en français reste peu étudié. Pour combler cette lacune, nous présentons HistoiresMorales, un jeu de données français dérivé de MoralStories, traduit puis affiné avec des locuteurs natifs pour assurer précision grammaticale et ajustement culturel. Afin de favoriser de futures recherches, nous menons des expériences préliminaires sur l’alignement des modèles multilingues en français et en anglais. Bien que les modèles de langage s’alignent généralement sur les normes morales humaines, nous observons qu’ils restent influençables, tant vers un alignement moral qu’immoral.

Aligning language models with human values is crucial, especially as they become more integrated into everyday life. While models are often adapted to user preferences, it is equally important to ensure they align with moral norms and behaviours in real-world social situations. Despite significant progress in languages like English and Chinese, French has seen little attention in this area, leaving a gap in understanding how LLMs handle moral reasoning in this language. To address this gap, we introduce HistoiresMorales, a French dataset derived from MoralStories, created through translation and subsequently refined with the assistance of native speakers to guarantee grammatical accuracy and adaptation to the French cultural context. We also rely on annotations of the moral values within the dataset to ensure their alignment with French norms. HistoiresMorales covers a wide range of social situations, including differences in tipping practices, expressions of honesty in relationships, and responsibilities toward animals. To foster future research, we also conduct preliminary experiments on the alignment of multilingual models on French and English data and the robustness of the alignment. We find that while LLMs are generally aligned with human moral norms by default, they can be easily influenced with user-preference optimization for both moral and immoral data.

Recent studies introduced effective compression techniques for Large Language Models (LLMs) via post-training quantization or low-bit weight representation. Although quantized weights offer storage efficiency and allow for faster inference, existing works have indicated that quantization might compromise performance and exacerbate biases in LLMs.This study investigates the confidence and calibration of quantized models, considering factors such as language model type and scale as contributors to quantization loss.Firstly, we reveal that quantization with GPTQ to 4-bit results in a decrease in confidence regarding true labels, with varying impacts observed among different language models. Secondly, we observe fluctuations in the impact on confidence across different scales. Finally, we propose an explanation for quantization loss based on confidence levels, indicating that quantization disproportionately affects samples where the full model exhibited low confidence levels in the first place.We make our code and quantized models publicly available.