# AAELI

Large language models can label every instance cheaply, but their judgments are biased and their errors correlated, so treating them as human labels can mislead statistical inference. AAELI uses model outputs as auxiliary information to guide a limited human audit and builds anytime-valid confidence sequences from predictable empirical-likelihood processes, so coverage survives data-dependent stopping. On Chatbot Arena, AAELI needs about 870 human labels on average to reach a confidence-sequence width of 0.10, versus about 1,100 on average for the strongest baseline, RASI, a saving of roughly 21%.

Large language models can label every instance cheaply, but their judgments are biased and their errors correlated, so treating them as human labels can mislead statistical inference. AAELI uses model outputs as auxiliary information to guide a limited human audit and builds anytime-valid confidence sequences from predictable empirical-likelihood processes, so coverage survives data-dependent stopping. On Chatbot Arena, AAELI needs about 870 human labels on average to reach a confidence-sequence width of 0.10, versus about 1,100 on average for the strongest baseline, RASI, a saving of roughly 21%.