I think when you mentioned the world of plastic deformation in your statement, the answer became very clear because the fact that spring effect is closely related to the elastic properties, which means almost nil hysteresis and system turns into original configuration after releasing the load (in your case torque). But if your opposite torque is not very large the spring returns its original state shape and oscillate back and fort till it dissipates stored elastic strain energy by frictional forces associated with the environment (assumed that both ends are exposed clamped boundary and motion takes place along the axial direction).

Note: There is critical upper limit for the pitch height under revers torque above which unfolding might take place if the stiffness is too small such as the case for alpha-peptide (ssDNA). (Stiffness is proportional with the fourth power of diameter for the usual torque term otherwise the second order. SEE our Paper in RG )

Thank you Tarık Ömer Oğurtanı