Moving samples toward high-reward regions instead of reweighting them
One post praises this approach, arguing that moving samples makes more use of the sample spaceβs geometric structure than reweighting does.
TLDR
The post highlights work that transports samples toward high-reward regions rather than reweighting them. It praises the approach for making more use of the geometric structure of the sample space.
Combined views
361
1 Source, first seen 1h ago
1 likes