Applying factor analysis to Hugging Face leaderboard benchmarks yields a different, arguably better LLM ranking than the simple average, with an inverted U-shaped relationship between the two.
1 The initial unidimensional model calibrated on the older leaderboard dataset, exhibited relatively poor model fit under Maximum Likelihood Robust estimator (Yuan & Bentler, 2000)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
Improving LLM Leaderboards with Psychometrical Methodology
Applying factor analysis to Hugging Face leaderboard benchmarks yields a different, arguably better LLM ranking than the simple average, with an inverted U-shaped relationship between the two.