A Fisher random walk weighted residual estimator achieves semiparametric efficient confidence intervals for contextual Bradley-Terry-Luce preference comparisons with flexible score estimators.
Lecture Notes on Spectral Graph Methods
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
These are lecture notes that are based on the lectures from a class I taught on the topic of Spectral Graph Methods at UC Berkeley during the Spring 2015 semester.
fields
stat.ML 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Fisher Random Walk: Automatic Debiasing Contextual Preference Inference for Large Language Model Evaluation
A Fisher random walk weighted residual estimator achieves semiparametric efficient confidence intervals for contextual Bradley-Terry-Luce preference comparisons with flexible score estimators.