Can vector-like lepton couplings to different lepton flavors be individually resolved, or do they always correlate in global fits?

# Resolving Vector-Like Lepton Couplings to Different Flavors

The short answer: they can be individually resolved, but only with complementary multi-channel measurements — they're strongly correlated in any single global fit unless you break the degeneracy with flavor-specific data.

Why correlations arise

In global fits of vector-like leptons, different flavor couplings (electron, muon, tau) tend to correlate because they often feed into the same observables through loop diagrams and mixing effects. A global fit that treats all parameters simultaneously can easily produce flat directions — regions of parameter space that give equally good fits to data [AG-2024.03-1124].

The issue is particularly acute for semileptonic operators and four-fermion interactions, where "several flat or at least very poorly constrained directions" prevent full determination of the parameter space [AG-2024.03-1124]. This means that without additional constraints, you cannot uniquely pin down, say, the coupling strength to electrons versus muons.

Breaking the degeneracy: flavor-specific observables

The key to resolving individual couplings is to use flavor-violating and flavor-specific observables that directly probe different lepton generations:

• Higgs decays to specific lepton pairs (H → ee, H → μμ, H → ττ) constrain each flavor separately.
• Rare decays like μ → eγ, τ → μγ, and μ-e conversion in nuclei are highly sensitive to flavor-off-diagonal couplings [AG-2026.03-1618].
• Lepton dipole moments (muon g−2, electron g−2) offer complementary handles on flavor-universal and flavor-specific contributions [AG-2024.03-1164].

One comprehensive analysis [AG-2026.03-1618] surveys "robust and complementary probes that allow to distinguish between models" by systematically combining Higgs decays, Z decays, rare lepton processes, and collider signatures. This multi-channel approach breaks degeneracies that would persist in any single observable.

Why this matters

Without this separation, claiming a discovery of vector-like leptons risks fitting to a degenerate solution where the model looks good on average but the individual flavor couplings are unphysical. Future precision experiments—MEGII (μ → eγ), Mu2e/COMET (μ-e conversion), and Belle II—will dramatically improve the flavor separation [AG-2026.03-1618].

---

Note: The provided preprints address flavor correlations in SMEFT operators and lepton flavor violation broadly, but don't directly discuss vector-like lepton flavor resolution in parametric detail.