<p>The Yellow River Basin has gathered a large number of energy and heavy chemical enterprises, and its carbon emissions account for more than one-third of the national total. Studying the issue of carbon emission efficiency in this region holds significant practical importance. The article constructs an ultra-efficiency model based on undesired output to measure the carbon emission efficiency of the Yellow River Basin. At the same time, it uses the modified gravity model and social network analysis method to explore the spatial network centrality characteristics. Additionally, it employs the quadratic assignment program regression analysis method to analyze the different factors influencing efficiency. Through empirical research, it is found that: (1) The overall carbon emission efficiency of cities in the Yellow River Basin shows a distribution pattern of "higher in the east and lower in the west, with provincial capitals leading and resource-based cities under pressure". (2) The overall density of the regional carbon emission network is stable, but locally it presents a core-periphery structure. (3) The impact of science and technology on the regional carbon emission network is manifested as a significant positive effect. Therefore, constructing a collaborative emission reduction system, strengthening network governance, and optimizing policy guarantees are conducive to promoting the improvement of carbon emission efficiency.</p>