Miller-Xu

为啥模型中的geo_mask是大于阈值设置为1 ```python sh_mx = sh_mx.T self.geo_mask = torch.zeros(self.num_nodes, self.num_nodes).to(self.device) self.geo_mask[sh_mx >= self.far_mask_delta] = 1 self.geo_mask = self.geo_mask.bool() ``` 原文写的是 From the short-range view, we define the binary geo-graphic masking matrix...