S4E2: Google DeepMind’s Dr. Claire Cui on The Next Frontier for Large Language Models
Theory and Practice
The Future of Uncertainty Quantification
I think there are a few gaps in factuality, grounding and introspection. These are related to one big area of gap that we have right now, LM. And I think another gap or opportunity is efficiency. If you want to be capable and want to have like say high quality, then we require hundreds of billions of weights fully connected,. That's just not sustainable for the longer term. Sparse modeling and this dynamic decision of which pathway to trigger or activate is going to be an important area to lead to really human level"
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
