Wang Yan: Filling the thinking chain moat of proprietary models

Wang Yan trained the model through native thinking chain training, making its performance surpass the proprietary model Claude's Fable 5 and GPT-6.1-SOL. The native thinking chain model performs excellently in benchmark tests, successfully bypassing the thinking chain moat of proprietary models. This technology provides users with an engineering solution to bypass the limitations of proprietary models.