Bias in Word Embeddings
Emily discusses how word embeddings can inadvertently incorporate societal biases, particularly in the context of Mexican restaurants and negative sentiment. She emphasizes the importance of using curated data to mitigate these biases, acknowledging that while it's impossible to eliminate bias entirely, improvements can be made. The conversation highlights the critical need for awareness in the use of external datasets in machine learning models.In this clip
From this podcast
Gradient Dissent - A Machine Learning Podcast
Emily M. Bender — Language Models and Linguistics
Related Questions
What role do biases play in learning as discussed in the episode Language Understanding and LLMs with Christopher Manning - 686 and the clip Learning with Bias?
How do vector embeddings work in the context of the episode Mindscape 280 | François Chollet on Deep Learning and the Meaning of Intelligence and the clip Understanding Language Models?
How do vector embeddings work in the context of the episode Mindscape 280 | François Chollet on Deep Learning and the Meaning of Intelligence and the clip Understanding Language Models?