List your company

Question Answering on BBQ

92.12AccuracyFairSteer
29.813645.989362.16578.3407Apr 20, 2025

Evaluation results

MethodLinks
FairSteerBackbone=Llama3-8B, Setting=Few-shot2025.0492.124.392.32——————
FairSteerBackbone=Llama3-8B, Setting=Zero-shot2025.0490.221.462.17——————
FairSteerBackbone=Vicuna-13B, Setting=Few-shot2025.0486.561.284.33——————
Llama3-8BSetting=Few-shot2025.0484.7413.532.42——————
CALBackbone=Llama3-8B, Setting=Few-shot2025.0482.652.612.69——————
CALBackbone=Mistral-7B, Setting=Few-shot2025.0481.898.434.68——————
FairSteerBackbone=Llama2-13B, Setting=Few-shot2025.0480.261.583.68——————
FairSteerBackbone=Mistral-7B, Setting=Few-shot2025.0479.938.044.87——————
FairSteerBackbone=Vicuna-13B, Setting=Zero-shot2025.0477.740.12.5——————
FairSteerBackbone=Llama2-13B, Setting=Zero-shot2025.0474.02-0.820.84——————
FairSteerBackbone=Mistral-7B, Setting=Zero-shot2025.0473.433.755.02——————
FairSteerBackbone=Vicuna-7B, Setting=Few-shot2025.0471.282.87.8——————
Llama3-8BSetting=Zero-shot2025.047113.622.51——————
Mistral-7BSetting=Few-shot2025.0470.5316.685.1——————
CALBackbone=Mistral-7B, Setting=Zero-shot2025.0468.322.284——————
FairSteerBackbone=Vicuna-7B, Setting=Zero-shot2025.0465.381.475.16——————
Vicuna-13BSetting=Few-shot2025.0464.7415.725.49——————
CALBackbone=Vicuna-13B, Setting=Few-shot2025.0463.7212.115.93——————
Vicuna-13BSetting=Zero-shot2025.0463.714.973.56——————
Mistral-7BSetting=Zero-shot2025.0462.2210.35.24——————
CALBackbone=Llama3-8B, Setting=Zero-shot2025.0455.510.085.64——————
FairSteerBackbone=Llama2-7B, Setting=Few-shot2025.0453.342.634.55——————
CALBackbone=Llama2-13B, Setting=Few-shot2025.0453.279.825.47——————
CALBackbone=Llama2-13B, Setting=Zero-shot2025.0451.291.412.46——————
Llama2-13BSetting=Zero-shot2025.0448.65.862.91——————
CALBackbone=Vicuna-13B, Setting=Zero-shot2025.0447.990.721.23——————
Llama2-13BSetting=Few-shot2025.0447.9416.315.55——————
FairSteerBackbone=Llama2-7B, Setting=Zero-shot2025.0446.281.042.11——————
Vicuna-7BSetting=Few-shot2025.0443.8914.289.07——————
Llama2-7BSetting=Few-shot2025.0442.676.435.36——————
CALBackbone=Llama2-7B, Setting=Zero-shot2025.0441.40.871.62——————
Vicuna-7BSetting=Zero-shot2025.0441.336.785.94——————
CALBackbone=Vicuna-7B, Setting=Few-shot2025.0440.3415.339.73——————
CALBackbone=Llama2-7B, Setting=Few-shot2025.0438.756.236.99——————
CALBackbone=Vicuna-7B, Setting=Zero-shot2025.0433.45-0.01-0.02——————
Llama2-7BSetting=Zero-shot2025.0432.212.442.41——————
Direct-Prompting StandardBase Model=gpt-3.5-turbo, Prompting Strategy=Direct-Prompting (Standard)2026.01———86.8264.2475.530.06360.05170.0577
MADBase Model=gpt-3.5-turbo2026.01———92.9241.1667.040.00680.05280.0298
MPTBase Model=gpt-3.5-turbo2026.01———85.0664.9274.990.01150.01660.0141
Re-Prompting DebiasBase Model=gpt-3.5-turbo, Prompting Strategy=Re-Prompting (Debias)2026.01———96.3655.2575.80.02730.01320.0203