Core Viewpoint - A recent study published on GitHub has sparked significant discussion in the industry, suggesting that Huawei's Pangu Pro MoE model and Alibaba's Qwen-2.514B model exhibit "astonishing similarities" in their parameter structures [1] Group 1: Model Comparison - The study found that the average correlation in attention parameter distribution between Pangu Pro MoE and Qwen-2.514B is as high as 0.927, which is significantly above the normal range observed in other model comparisons [1] - This high correlation has led to speculation among netizens that Pangu Pro MoE may have engaged in plagiarism [1] Group 2: Responses and Reactions - The Pangu Pro MoE team has responded on GitHub, denying the plagiarism allegations and asserting that the author's evaluation method is unscientific [1] - As of the time of reporting, Huawei has not provided an official comment on the matter [1]