How To Find Similar Benchmarks Or Datasets On Huggingface?

Recently I’m stuck hard on my research at finding extremely similar datasets, but should still be different. It seems that even datasets in the same semantic field differs drastically inside an LLM’s internal representations. Any good way of finding datasets this similar?

By the way I’m especially interested in datasets that are at least 5,000 samples large, which narrows the candidates greatly.