List your company

Question Answering on HH-RLHF

59AccuracyMA-PoP
43.81647.75851.755.642May 7, 2026

Evaluation results

MethodLinks
MA-PoP2026.0559
Decentr. MADT=22026.0558.67
Decentr. MADT=32026.0558.67
Sparse MADT=22026.0557.33
Self-Consistency2026.0557.33
Decentr. MADT=52026.0557
Free MADT=52026.0556.67
ISP2026.0556.67
Majority Voting2026.0556.33
Sparse MADT=52026.0556
Mistral-7B2026.0555.67
Gemma-9B2026.0555.67
Centr. MADT=32026.0555.67
Free MADT=32026.0555.67
Centr. MADT=22026.0555.33
Sparse MADT=32026.0555.33
Free MADT=22026.0555
Centr. MADT=52026.0553.67
LLP2026.0553.67
Falcon-7B2026.0550.33
Qwen-7B2026.0547.73
Llama-8B2026.0544.4