Tag
The paper proposes a novel framework (CDDTLDA) using transfer learning and data augmentation to improve Chinese dialects discrimination under low-resource conditions, achieving state-of-the-art results on two benchmark corpora.
Amnesty International's briefing argues that generative AI systems built on unlawful web scraping violate international human rights law, and calls for their prohibition.
This essay argues that evaluation is the hardest problem in production AI, not generation, and decomposes AI self-knowledge into calibration, discrimination, and expression, with implications for system design.
ArabDiscrim is a decade-long lexical resource and corpus of 293K Arabic Facebook posts about racism and discrimination, with engagement signals, morphological regex families, and discrimination axes, supporting fairness-oriented Arabic NLP research.