Tag
Introduces CuBAS, an information-geometric framework for adaptive data selection in supervised classification that uses local curvature of the data manifold to identify informative samples, achieving improved accuracy across 30 benchmark datasets.
This paper formulates adaptive sampling for large language models as a Markov decision process and trains a lightweight RL controller to balance correctness, latency, and computational cost, achieving improved trade-offs.
The NOVA framework models the 'generate, verify, accumulate, retrain' loop as an adaptive sampling process over a knowledge space, identifying failure modes and proving a scaling law for cumulative generation cost under Zipf-like discovery distributions.