Surprise-Guided MergeSort uses a VLM-based composite surprise scorer to prioritize human comparisons inside a MergeSort scheduler, skipping up to 535 pairs per session and raising Kendall's τ by 6-12 points over Active Elo at fixed budget across six benchmarks.
arXiv preprint arXiv:2202.04823 (2022)
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
MetaRanker uses active learning with human preference judgments and lightweight VLM priors to rank metalens images by semantic interpretability, achieving closer human alignment with roughly 80% fewer pairwise annotations than exhaustive comparison.
citing papers explorer
-
MetaRanker: Human-in-the-loop Active Ranking for Metalens Image Quality
MetaRanker uses active learning with human preference judgments and lightweight VLM priors to rank metalens images by semantic interpretability, achieving closer human alignment with roughly 80% fewer pairwise annotations than exhaustive comparison.