Research & Resources
Our work is informed by leading research in the field of responsible AI.
Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings
Seminal work identifying gender stereotypes in word embeddings and proposing methods to debias them.
Detecting and Mitigating Harmful Outputs in Generative AI
Techniques to identify and reduce harmful content generated by advanced AI models.
Bias in Multimodal AI Systems
Analysis of bias propagation in multimodal AI systems combining text, image, and audio data.