BenchBuilder automates extraction of 500 challenging prompts from crowd-sourced data into Arena-Hard-Auto, delivering 3x model separation versus MT-Bench and 98.6% human-preference correlation at $20 cost.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipeline
BenchBuilder automates extraction of 500 challenging prompts from crowd-sourced data into Arena-Hard-Auto, delivering 3x model separation versus MT-Bench and 98.6% human-preference correlation at $20 cost.