Share your thoughts, 1 month free Claude Pro on usSee more

Fairness Evaluation on CrowS-Pairs

72.2Score

Self-Debias Iter2 + Self-Correction

Updated 8d ago

Evaluation Results

Method	Links
Self-Debias Iter2 + Self-Correction 2026.04		72.2
Self-Debias Iter2 2026.04		71.2
Self-Debias Iter1 + Self-Correction 2026.04		70.2
Self-Debias Iter1 2026.04		70
Qwen1.5-8B 2026.04		68.8
Qwen1.5-8B + Self-Correction 2026.04		68.8
Self-Debias Offline + Self-Correction 2026.04		68.5
Self-Debias SFT 2026.04		68.2
Self-Debias Offline 2026.04		67.8
Self-Debias SFT + Self-Correction 2026.04		67.5
Qwen2.5-7B-Instruct 2026.04		66.5
DeepSeek-R1-Distill-Qwen-7B 2026.04		59.2
Qwen2.5-7B-Instruct + Self-Correction 2026.04		59.2
DeepSeek-R1-Distill-Qwen-7B + Self-Correction 2026.04		58.5
Llama-3.1-8B-Instruct 2026.04		54.2
Llama-3.1-8B-Instruct + Self-Correction 2026.04		51