Model Forgetting Challenges
The conversation delves into the complexities of making AI models "forget" specific training data, with a cautionary note about the feasibility of such an endeavor. Emphasis is placed on the growing availability of open-source models that provide transparency through weights and biases reports, allowing users to investigate the training data used. This transparency is particularly relevant in light of ongoing copyright discussions in the AI field.In this clip
From this podcast
Eye on AI
How Weights and Biases Supercharges Machine Learning | Lukas Biewald
Related Questions