Datasets for reproducing the results of [Reinforcement Learning for Adaptive Composition of Quantum Circuit Optimisation Passes] (arxiv/2601.21629)

This listing contains datasets for reproducing the results of [Reinforcement Learning for Adaptive Composition of Quantum Circuit Optimisation Passes] (https://arxiv.org/abs/2601.21629). Included are the following .zip files which contain the .csv files that can be used to generate the plots in the paper:in_distribution_seed*.zip - .csv files containing results for 'in-distribution' testing for a given model corresponding to a given training seed. out_of_distribution.zip - .csv files containing results for 'out-of-distribution' testing as part of the 'generalisability' section. very_large_circuits.zip - .csv files containing results for 'very large out-of-distribution' testing as part of the 'generalisability' section. beam_search.zip - .csv files containing results for the 'beam-search comparison' as part of the 'scalability' section (different structure to the other files; each row shows a compiled circuit, including its original ID, optimisation method, reward, runtime, 2-qubit gate counts before/after, and applied pass sequence). The following .zip files contain the training/testing data used in the paper, all data in .json format and containing a 'data.pkl' DataFrame file describing the circuits:iqp_qaoa_rand_pauli_long_ordered_kak_cliff_squash_small.zip - training and testing data. iqp_qaoa_rand_pauli_long_ordered_kak_cliff_squash_small_testing_ood.zip - testing data for 'out-of-distribution' data as part of the 'generalisability' section. very_large_circuits_testing.zip - testing data for 'very large out-of-distribution' data as part of the 'generalisability' section.beam_search_test_circuits.zip - testing data for the 'beam-search comparison' as part of the 'scalability' section.