{"url":"/dataset/elements","name":"Elements","full_name":null,"description_markdown":"A configurable synthetic dataset of simple shapes with ground truth concepts and known causal relationships between concepts and classes.\r\n\r\nThe dataset does not consist of a specific set of images, instead it is provided as code to generate images matching the dataset. The specific combination of shapes, colours, textures and number of objects in each image is configurable, along with the definitions of each class. \r\n\r\nThe dataset is intended to be used to help study concept-based interpretability methods as it gives you full control over the concepts that have a causal influence on the class and over the associations between concepts within the dataset.","description_withheld":null,"homepage":"https://github.com/AngusNicolson/elements","introduced_date":"2024-04-04","introduced_date_note":null,"introduced_by":{"paper":"/paper/explaining-explainability-understanding","title":"Explaining Explainability: Recommendations for Effective Use of Concept Activation Vectors","first_author":"Angus Nicolson","url":null},"license":{"name":"MIT","url":null},"modalities":[{"name":"Images","url":"/datasets/modality/images"}],"tasks":[],"languages":[],"variants":["Elements"],"data_loaders":[],"num_papers_in_archive":1,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[],"papers_with_a_benchmark_row":[],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":0,"samples_harvested":0,"samples_ran":0,"samples_unverified":0,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}