List your company

Multilingual Question Answering on mGPQA

32.9Accuracyself-cons (ours)
18.23622.04325.8529.657May 31, 2026

Evaluation results

MethodLinks
self-cons (ours)Model Size=14B2026.0532.9
S1Model Size=14B2026.0532.1
LIDRModel Size=14B2026.0529.4
self-cons (ours)Model Size=7B2026.0528.9
baseModel Size=14B2026.0528.2
S1Model Size=7B2026.0525.7
self-cons (ours)Model Size=1.5B2026.0524.9
baseModel Size=7B2026.0524.6
S1Model Size=1.5B2026.0524.5
LIDRModel Size=7B2026.0523.5
baseModel Size=1.5B2026.0522.8
LIDRModel Size=1.5B2026.0518.8