Navigating Computational Trade-offs
Daniel and Peter discuss the shift in expectations for DPARL papers to include multiple runs and the trade-offs involved in determining the necessary number of runs. They also highlight the rise of purely empirical papers and the importance of analyzing algorithm performance for real-world deployment.In this clip
From this podcast
The Gradient
Peter Henderson on RL Benchmarking, Climate Impacts of AI, and AI for Law
Related Questions