Probing SMEFT operators through t t ‾ t t ‾ production with hypergraph neural networks at the LHC

We present a phenomenological study of tt̄tt̄ production in proton-proton collisions at √s = 13~TeV, using a Hyper-Graph Neural Network (H-GNN) to discriminate multilepton signal events from the dominant SM backgrounds, namely tt̄W, tt̄Z, tt̄H, tt̄VV, single-top associated production, and diboson and triboson processes. In the H-GNN architecture each event is represented as a hypergraph whose nodes correspond to reconstructed jets and leptons and whose hyperedges encode higher-order correlations among arbitrary subsets of these objects, allowing the network to learn the many-body kinematic structures that characterize the tt̄tt̄ final state. Combining same-sign di-lepton, tri-lepton, and four-lepton channels following a CMS-like event selection, the H-GNN attains an area under the ROC curve of $0.951$ for the tt̄tt̄ signal and yields a statistical significance of $Z = 9.11$ at an integrated luminosity of L = 140~fb⁻¹, to be compared with $Z = 8.62$ for a SPANet baseline, $Z = 7.37$ for a Particle Transformer baseline, and $Z = 5.13$ obtained by the ATLAS analysis, evaluated under identical event selection. We exploit the improved signal extraction to derive one- and two-parameter $95%$ confidence level limits on the Wilson coefficients of the dimension-six operators OΦu, O⁽¹⁾ₜₜ, O⁽¹⁾qq, O⁽¹⁾qt, and O⁽⁸⁾qt, and we project the expected sensitivity at the HL-LHC integrated luminosities of 1000~fb⁻¹ and 3000~fb⁻¹ with $50%$ uncertainty on the background estimation.