List your company

Scientific Reasoning on UGPhysics English

33.1Overall AccuracySci-CoE-Stage 2
13.922418.901223.8828.8588Feb 12, 2026

Evaluation results

MethodLinks
Sci-CoE-Stage 2Backbone=Qwen3-8B, Data Scale=30k2026.0233.130.5134.834.99
Sci-CoE-Stage 2Backbone=Qwen3-8B, Data Scale=18k2026.0232.4630.2133.334.38
Sci-CoE-Stage 1Backbone=Qwen3-8B, Data Scale=4k2026.0232.0330.2530.6234.38
Qwen3-8B Base ModelData Scale=None2026.0231.7630.7329.9833.51
Mistral-Small-Instruct-2409Data Scale=None2026.0225.7222.7122.729.97
Sci-CoE-Stage 2Backbone=Qwen2.5-7B-Instruct, Data Scale=30k2026.0222.6421.8423.1324.91
Sci-CoE-Stage 2Backbone=Qwen2.5-7B-Instruct, Data Scale=18k2026.0221.9220.9221.3123.17
Sci-CoE-Stage 1Backbone=Qwen2.5-7B-Instruct, Data Scale=4k2026.0221.0720.1419.8122.51
Qwen2.5-7B-Instruct Base ModelData Scale=None2026.0220.6718.8818.5223.34
Yi-1.5-9B-ChatData Scale=None2026.0217.611615.8519.94
Mathstral-7B-v0.1Data Scale=None2026.0217.4514.8217.7719.94
Ministral-8B-Instruct-2410Data Scale=None2026.0216.3913.9515.5219.2
Llama-3.1-8B-InstructData Scale=None2026.0214.6612.6414.3516.8