If vector-like leptons modify both loops and tree decays, how do global fits disentangle their individual contributions?

Question

Antigravity · Accepted Answer

This is an excellent question about a central challenge in precision physics. Let me walk through how global fits disentangle loop and tree contributions from vector like leptons. The Core Problem Vector like leptons (VLLs) contribute to observables in two fundamentally different ways: at tree level (direct exchange) and at one loop (virtual particle running in diagrams). Since both modify the same observable—say, a rare decay branching ratio or a precision measurement—naive single observable constraints can't tell them apart. A global fit solves this by exploiting correlations across many observables. How Global Fits Exploit Correlations The key insight is that tree and loop contributions have different "fingerprints" across the observable landscape. Here's the strategy: 1. Operator mixing generates distinct patterns [AG 2024.08 976] When you run effective field theory equations down from high energy (where new physics lives) to low energy (where you measure things), tree level operators mix with loop level operators through renormalization group equations. This coupling creates correlated predictions: if a VLL contributes to process A at tree level, it automatically predicts specific loop contributions to process B. A global fit checks whether all these predictions hang together. 2. Lepton flavor and process dependence Vector like leptons modify both lepton flavor conserving and lepton flavor violating channels, often with different coupling structures [AG 2026.03 1618]. For example, they affect Higgs decays, Z decays, muon g−2, and rare processes like μ→eγ simultaneously. Loop contributions to dipole moments scale differently with coupling strength than tree contributions to leptonic decays [AG 2024.08 1174]. By fitting all channels together, you over constrain the model: a parameter set that fits one observable's loop contribution may predict a tree level branching ratio that contradicts another measurement. 3. Flavor structure and semileptonic constraints In models like the $U 1$ vector leptoquark, the leptoquark couples to specific quark lepton pairs. This means tree level contributions to charged B decays are linked to loop corrections in the B decay constant and flavor changing neutral currents [AG 2025.11 1220]. A global fit enforces consistency: the same coupling that explains a deviation in $B 	o \ell 
u$ must also satisfy constraints from related processes like $B s 	o \mu^+ \mu^ $ [AG 2025.10 1692]. Practical Example: Flattening Directions When you fit to a single observable with both tree and loop parameters free, you often find "flat directions"—combinations of parameters that all predict the same result, leaving you unable to disentangle them. Global fits reduce (or eliminate) these degeneracies by adding constraints from processes that depend on the parameters differently [AG 2024.03 1124]. For instance, a tree contribution weighted by one Clebsch–Gordan coefficient enters one decay, while a loop contribution with different loop integrals enters another; fitting both simultaneously pins down the actual parameter values. Systematic Coverage A comprehensive study maps out the allowed parameter space across multiple fronts: collider bounds (LHC Higgs and Z decays), precision measurements (muon g−2, electron dipole moment), and rare decays [AG 2026.03 1618]. The interplay matters: loop induced dipole moments constrain chirality flipping couplings, while tree level semileptonic decays constrain vector couplings—and these are related through the VLL mass and mixing angles. A global fit enforces all these relationships at once. The preprints focus heavily on specific models (vector leptoquarks, vector like leptons in SMEFT) and individual processes rather than the meta question of how a global fit systematically disentangles loop vs. tree effects across a portfolio of observables.