Correct treatment of the broken-phase saddle in 4D φ^4 theory

Determine the correct non-perturbative treatment of the broken-phase saddle solution in four-dimensional scalar φ^4 theory, specifying the appropriate relations between broken-phase bare and renormalized parameters (such as choices for \(\tilde\lambda_B\), \(\tilde m_B^2\), and \(\tilde\Lambda_{\overline{\rm MS}}\)) and their identification relative to the symmetric-phase parameters, so that the renormalized free energy and pole mass are consistently defined and comparable under the sign-flip coupling relation \(\lambda_B=-2\tilde \lambda_B\).

Background

The paper constructs interacting saddle point expansions (R1-level resummations) for scalar φ^4 theory in both symmetric and broken phases. In dimensions d=2 and d=3, these analyses reproduce known dualities (Chang and Magruder) and qualitatively recover the phase structure. In d=4, after renormalization, the symmetric and broken-phase free energies and pole masses coincide under a sign-flip of the coupling, suggesting an exact self-duality.

However, the author highlights an unresolved issue regarding how to correctly treat the broken-phase saddle in d=4. Several prescriptions are proposed, including matching bare mass parameters ($\tilde m_B^2=m_B^2$), enforcing identical bare couplings ($\tilde \lambda_B=\lambda_B$), or adjusting the renormalization scales (e.g., $\tilde \Lambda_{\overline{\rm MS}}$ related to $\Lambda_{\overline{\rm MS}}$ via $\bar\mu$). Clarifying the correct treatment is necessary to fix the broken-phase saddle consistently and to interpret the proposed self-duality in four dimensions.