This release includes a dataset generated using the data-generation mechanics provided in the repository. The dataset was created with the seed term "Computer science," a depth of 10, and a layer size of 1000.
Additionally, the release contains the optimized weights for the encoder and predictor models on the best performing configuration.