Accuracy Ratings
Learn about Genie AI's accuracy ratings and read our machine learning research paper on ‘model robustness’.
We know that evaluating the output of generative AI is challenging, and we take this very seriously.
We compute accuracy ratings from expert lawyer annotated data, and preference data we gather from our legal users.
In collaboration with Oxford University, we've published research on "model robustness" in deep neural networks. This research advances our understanding of AI consistency and accuracy and guides our development of more reliable legal AI systems.
Read our research paper on model robustness here.
Got any more questions? Please reach out to us. We're always happy to talk about accuracy.
