AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Future of Research in AI
It was actually motivated by biomedical problem, but it's widely applicable, broadly applicable to other applications. And then also this emerging models of, you know, LLMs or other, you know,. foundational models. How to how to compute the feature attributions properly is an open problem. We are really interested in sample based importance to say that you transpose the matrix transpose over your feature matrix. So I've been talking about this feature attributions a lot and we can also apply Shapley values to gain insights into which samples are important for model training.