List your company

Instruction Following on AlpacaEval (OOD)

24.9KL Div (α=1)GEB-arctanh(π − 1)
24.765225.675126.58527.4949Sep 27, 2025

Evaluation results

MethodLinks
GEB-arctanh(π − 1)Backbone=LLaMA-3-8B-SFT, Evaluation Protocol=Out-of-domain2025.0924.925.9619.6223.49
f-DPOBackbone=LLaMA-3-8B-SFT, Evaluation Protocol=Out-of-domain2025.0925.7224.7317.822.75
FEBBackbone=LLaMA-3-8B-SFT, Evaluation Protocol=Out-of-domain2025.0925.7223.7519.6223.03
GEB-1/πBackbone=LLaMA-3-8B-SFT, Evaluation Protocol=Out-of-domain2025.0926.125.2819.823.73
GEB-πBackbone=LLaMA-3-8B-SFT, Evaluation Protocol=Out-of-domain2025.0928.2725.8720.0524.73