Hit Rate@K (Hit Rate at Top-K) is an important evaluation metric for Top-K and session-based recommendation tasks in natural language processing. Hit Rate@K refers to the proportion of users whose ground-truth interacted item appears in the top K positions of the recommendation list, and it is the most basic metric for measuring the overall recall ability of a recommendation system. It is usually reported as a percentage or a score between 0 and 1. This metric is widely used in relevant benchmarks and papers, and it can effectively reflect a model's performance along this dimension. Higher values are better. It helps researchers and engineers quantify differences in actual model performance in Top-K and session-based recommendation scenarios, making it easier to compare different methods or model versions side by side.