List your company

Scientific Reasoning Domain Average

65.2AccuracyDASD
62.39263.12163.8564.579May 21, 2026

Evaluation results

MethodLinks
DASDBackbone=Qwen3-8B2026.0565.2
OPSDBackbone=Qwen3-8B2026.0564.3
GRPOBackbone=Qwen3-8B2026.0563
BaseBackbone=Qwen3-8B2026.0562.5