The fairness of Natural Language Processing (NLP) models has emerged as a crucial concern. Information theory indicates that to achieve fairness, a model should not be able to predict sensitive variables, such as gender, ethnicity, and age. However, information related to these variables often appears implicitly in language, posing a challenge in identifying and mitigating biases effectively. To tackle this issue, we present a novel approach that operates at the embedding level of an NLP model, independent of the specific architecture. Our method leverages insights from recent advances in XAI techniques and employs an embedding transformation to eliminate implicit information from a selected variable. By directly manipulating the embeddings in the final layer, our approach enables a seamless integration into existing models without requiring significant modifications or retraining. In evaluation, we show that the proposed post-hoc approach significantly reduces gender-related associations in NLP models while preserving the overall performance and functionality of the models. An implementation of our method is available: https://github.com/fanny-jourdan/TaCo

翻译：自然语言处理（NLP）模型的公平性已成为一个关键问题。信息论指出，为实现公平性，模型不应能预测敏感变量（如性别、种族和年龄）。然而，这些变量相关的信息常隐含在语言中，给有效识别和消除偏见带来挑战。针对此问题，我们提出一种新颖方法，该方法在NLP模型的嵌入层运行，且不依赖特定架构。我们的技术利用可解释人工智能（XAI）领域的最新进展，通过嵌入变换消除选定变量中的隐式信息。通过直接操作最终层的嵌入，该方法可无缝集成至现有模型，无需重大修改或重新训练。评估表明，所提出的后处理方法能显著降低NLP模型中的性别关联，同时保持模型的整体性能与功能。我们的方法实现已公开于：https://github.com/fanny-jourdan/TaCo