怎么评价现代拓扑凝聚态物理实验中关于“里程碑式发现”的证据及其可靠性问题?问题背景: DOI:10.1126/science.adk9181 Abstract Manipulating the topology of electronic bands can realize new states of matter, with possible implications for information technology. A central question is how to tell whether a topological regime has been achieved. Experiments are often guided by a prediction of a distinct and self-explanatory signal called “the smoking gun.” However, in micrometer- or nanometer-scale specimens, phenomenology can mimic the anticipated behavior without containing the exotic states. We show limited data that are consistent with the presence of four topological phenomena; by considering additional data, we identified the most likely origins of the observed patterns as trivial. We argue that the reliability of smoking gun–type claims can be greatly enhanced by releasing comprehensive datasets, discussing alternative scenarios, and disclosing the total volume of study. 在凝聚态物理实验中,尤其是在寻求拓扑状态和拓扑量子计算相关的实验中,研究人员常常希望看到某种明显的实验信号,这种信号被称为 smoking gun/决定性证据,号称能够证明达到了新的拓扑状态或证明某些重要的理论预测。 实际上,这些所谓 smoking gun 信号很可能并不独一无二。在微米或纳米尺度的复杂样本中,一些平凡的物理现象,例如微结构干涉、量子限制效应,在精细调整实验条件后可以产生和预期的“定性”信号相似的图案或特征,从而误导人们认为已经实现了拓扑状态。 过去的一些高影响力论文在尝试复现这些“突破性”信号时遇到了困难,一些复制工作因为“不够新颖”或“领域已向前发展”等原因而未能跟声称实现突破的文章那样在顶级期刊发表,这反映了当前实验文化在验证和复制方面的缺失。 论文作者强调数据共享和全面数据展示的重要性:如果研究团队能够公开更大体量、更全面的数据集,而不是只展示符合预期的局部结果,并且开放讨论可能的其他解释,那么整个领域就能更有效地避免错误的结论,从而提高研究的可靠性。 文章建议在同行评审和科学交流中更重视替代解释方案的探讨,而不仅仅是寻找单一的关键证据。