In this study, we employ a machine-learning potential approach based on first-principles calculations combined with the Boltzmann transport theory to investigate the impact of lithium-ion de-embedding on the thermal conductivity of LiFePO4, with the aim of enhancing heat dissipation in lithium-ion batteries. The findings reveal a significant decrease in thermal conductivity with increasing lithium-ion concentration due to the decrease in phonon lifetime. Moreover, removal of lithium ions from different sites at a given lithium-ion concentration leads to distinct thermal conductivities, attribu...