List your company
ResearchFormerly WizWand

Reward Modeling on PersonalLLM Near Uniform (α=0.1) Unseen

96.4AccuracyVRF
58.54468.37278.288.028Apr 1, 2026
Updated 7mo ago

Evaluation results

MethodLinks
VRF2026.0496.4
BT2026.0494.6
PAL2026.0492.8
LoRe2026.0492.5
VPL2026.0491.5
PReF2026.0490.4
Ref2026.0460