Binary Pronoun

Research on binary pronouns in natural language processing (NLP) focuses on mitigating gender bias in language models, particularly concerning their handling of gendered and gender-neutral pronouns. Current work investigates how different model architectures, such as encoder-based (e.g., BERT) and decoder-based (e.g., GPT) models, process contextual cues related to gender, revealing significant differences in their prioritization of information. This research is crucial for developing more equitable and inclusive NLP systems, impacting applications ranging from machine translation to coreference resolution, and addressing the misgendering and erasure of non-binary individuals.