§ feed · storyline
Interpretable machine learning through teaching
Researchers publish a method for interpretable machine learning that selects informative examples to teach concepts to both AI models and humans simultaneously.
We’ve designed a method that encourages AIs to teach each other with examples that also make sense to humans. Our approach automatically selects the most informative examples to teach a concept—for instance, the best images to describe the concept of dogs—and experimentally we found our approach to be effective at teaching both AIs
§ sources1 publication · timeline below