List your company

General Reasoning on Out-of-Domain Tasks (GSM, MATH, DivLogicEval, SocialIQA, CodeMMLU)

30.3GSM ScoreSOLAR
15.01218.98122.9526.919Mar 23, 2026

Evaluation results

MethodLinks
SOLARBase Model=Qwen2.5-0.5B-Instruct, Adaptation Method=SOLAR Level I2026.0330.324.525.15535.634.1
TTLBase Model=Qwen2.5-0.5B-Instruct2026.0323.519.726.234.929.726.8
DnDBase Model=Qwen2.5-0.5B-Instruct2026.0320.824.12133.529.125.7
DOMBase Model=Qwen2.5-0.5B-Instruct2026.0317.72.624.751.331.625.6
LoRABase Model=Qwen2.5-0.5B-Instruct2026.0315.66.820.339.529.822.4