Unfolding weights corresponding to a selection of jets from the Jet Primary Dataset of the CMS 2011A Open Data in MOD HDF5 format and associated simulated datasets. The unfolding is performed in a high-dimensional manner using the OmniFold method, which can unfold all observables simultaneously. Particle Flow Networks are used in Step 1 and Step 2 of the OmniFold method to process the full phase space information. The datasets and neural networks were accessed/built via the EnergyFlow Python package. An upcoming version of the package will contain an example/demo demonstrating how to use these weights. The phase space selections for the data, sim, and gen datasets (using the terminology of the OmniFold paper) are: data: \(p_T^{\rm jet}\in [375, 700]\) GeV, \(|\eta^{\rm jet}|<2.4\), jet quality \(\ge\) 2 sim: \(p_{T,\text{corr}}^{\rm jet} \in [375, 700]\) GeV, \(|\eta^{\rm jet}| < 2.4\), gen jet matched ('gen_jet_pts != -1' in EnergyFlow), jets from the 170 and 1800 MC datasets are excluded gen: Matched to sim jet The omnifold_weights.npz file contains two arrays, 'wssim' corresopnding to the Step 1 weights \(\omega_n\), and 'wsgen' corresponding to the Step 2 weights \(\nu_n\), for iteration \(n\). The shape of each of these arrays is (6, 16489054), with the first axis being the iteration axis and the second axis being the event axis. There are 5 iterations, but 6 sets of weights in each array, with the 0th entry being the starting weights.