Conceptadaptive sampling
Adaptive Sampling
An inference optimization method that dynamically adjusts the number of samples according to problem difficulty or model uncertainty, with the aim of balancing accuracy against computational cost.
Timeline (last 90 days)
Nov 6
Existing efficient reasoning variants (such as early stopping and adaptive sampling) mainly reduce cost from two angles—terminating entire trajectories early or adaptively allocating the number of samples—and do not examine the error distribution step by step within trajectories.
Unverified50%