Research

CLM-as-a-Judge: Evaluating an Open Contrastive Decision Model on Public Judge Benchmarks

Single source1 source7 Oct 2026

arXiv:2610.07177v1 Announce Type: cross Abstract: An open contrastive decision model is near chance as a judge on the hard public benchmarks: Contrastive-LM/CLM-v0.1-8B scores between 0.351 (best- of-four, chance 0.250) and 0.593 (pairwise, chance 0.500), is s

Reported by 1 source

arxiv.orghttps://arxiv.org/abs/2610.07177

← Back to all developments · Signals are a development reported by more than one source. See the methodology.