Operator Identification in Charged Lepton-Flavor Violation: Global EFT Analysis with RG Evolution, Polarization Observables, and Bayesian Model Discrimination at Future Colliders

This paper presents a global effective field theory analysis of charged lepton-flavor violation across multiple future colliders, incorporating renormalization-group evolution, polarization observables, and Bayesian model discrimination to achieve precise operator identification and UV physics discrimination beyond simple exclusion limits.

The Gist

Imagine the Standard Model of physics as a giant, incredibly detailed rulebook for how the universe works. For decades, this rulebook has been perfect at predicting almost everything we see. But physicists suspect there's a hidden chapter we haven't found yet—a "Beyond the Standard Model" (BSM) section that explains mysteries like dark matter or why the universe has more matter than antimatter.

This paper is about hunting for a specific, forbidden move in that rulebook: Charged Lepton Flavor Violation (CLFV).

The Forbidden Move: The "Flavor Swap"

In our current rulebook, particles called "leptons" (electrons, muons, and taus) have strict identities. An electron is always an electron; it never spontaneously turns into a muon. It's like a rule in a game of chess that says a Pawn can never suddenly become a Queen.

If we ever see an electron turn into a muon (or a muon into an electron) without a valid reason, it's a "null test" violation. It's the smoking gun that proves the rulebook is incomplete and that new, unknown physics is at play.

The Detective Work: Finding the Culprit

The authors of this paper aren't just looking for if this happens; they are trying to figure out how it happens.

Think of it like a crime scene.

• Old Approach: "We found a broken window! The criminal is inside!" (This is just finding the signal).
• This Paper's Approach: "We found a broken window, but was it a rock thrown by a kid, a bullet from a gun, or a ghost? We need to identify the tool used to break it."

In physics, these "tools" are mathematical operators (like dipoles, currents, or four-fermion interactions). The paper argues that simply finding the crime isn't enough; we need to identify the specific "weapon" to know which new theory of physics is correct.

The Toolkit: A Global Investigation

The authors built a massive, unified investigation plan that combines clues from three different types of "crime scenes" (colliders):

1. The Low-Energy Lab (The Quiet Room): Experiments like MEG or Mu3e look for rare decays in very quiet, controlled environments. They are like forensic experts looking for tiny fingerprints. They are great at finding specific clues but might miss the bigger picture.
2. The High-Energy Smashers (The Thunderdome): Machines like the Large Hadron Collider (LHC) or future colliders (FCC, Muon Colliders) smash particles together at incredible speeds. This is like a high-speed car crash investigation. It's messy, but it can reveal heavy, new particles that the quiet labs can't see.
3. The Future Machines: They looked ahead to machines that don't exist yet (like a 10 TeV Muon Collider) to see what they could teach us.

The Secret Sauce: How They Solve It

The paper introduces three clever tricks to solve the mystery:

1. The "Tomographic" View (Taking a 3D X-Ray)
Usually, scientists look at one angle at a time. This paper takes a "tomographic" approach. Imagine trying to understand a complex sculpture. If you only look at the shadow on the wall, you might think it's a flat circle. If you look from the side, it looks like a square.
The authors combine data from all angles (different colliders, different energies) to build a 3D model of the "forbidden move." This helps them see the true shape of the new physics, rather than just a flat shadow.

2. The "Spin" Trick (Polarization)
Imagine trying to tell the difference between a left-handed and a right-handed glove. If you just look at a pile of gloves, it's hard. But if you have a machine that can only catch left-handed gloves, you instantly know what you're dealing with.
The paper uses "beam polarization" (controlling the spin of the particles in the collider) to act like that filter. By turning the spin "left" or "right," they can separate different types of new physics that would otherwise look identical.

3. The "Bayesian" Scorecard (The Detective's Intuition)
Instead of just saying "We found a 5-sigma signal" (which means "we are very sure this happened"), they use a statistical tool called Bayes Factors.
Think of this as a detective's scorecard.

• Hypothesis A: The criminal is a Leptoquark (a specific type of new particle).
• Hypothesis B: The criminal is a Heavy Neutral Lepton.
  The paper calculates a score: "Given all the evidence, Hypothesis A is 5 times more likely than Hypothesis B." This helps them rule out entire families of theories, not just individual numbers.

The Results: What Did They Find?

• The "Mixing" Effect: They found that as you look at higher energies, the "tools" (operators) start to mix together, like colors of paint. If you ignore this mixing (which many older studies did), you get the wrong answer. They corrected for this, showing that the "tools" change slightly as you zoom in.
• The Power of Combination: A single experiment might be confused by two different theories looking the same. But when you combine the "Quiet Room" (low energy) data with the "Thunderdome" (high energy) data, the confusion disappears. The two theories look very different when viewed from both angles.
• The Future Plan: They created a "run-plan optimizer." This is like a GPS for future colliders. It tells scientists: "Don't just run the machine at full speed; run it with this specific mix of beam spins and energy levels to get the best chance of identifying the new physics."

The Bottom Line

This paper is a roadmap for the next generation of physics. It moves us from asking "Is there new physics?" to "What exactly is the new physics?"

By combining data from every possible angle, using advanced statistics, and accounting for how particles behave at different energies, they have built a system that can distinguish between different "suspects" in the universe's hidden sector. It's a shift from just finding a footprint to identifying the shoe, the brand, and the person who wore it.

Technical Summary

1. Problem Statement

Charged Lepton-Flavor Violation (cLFV) is a null-test of the Standard Model (SM); any observation would be direct evidence of Beyond-the-Standard-Model (BSM) physics. While low-energy experiments (e.g., MEG II, Mu3e, COMET) provide stringent constraints, and future colliders (FCC-ee, ILC, HL-LHC, Muon Colliders) offer high-energy probes, existing analyses often focus on exclusion reach (setting limits on Wilson coefficients) rather than operator identification.

The core problem addressed is the degeneracy in the Effective Field Theory (EFT) parameter space. Different combinations of Wilson coefficients can produce similar total event yields, making it difficult to distinguish between specific BSM models (e.g., Leptoquarks vs. Heavy Neutral Leptons). The paper argues that the primary goal of future collider programs should shift from merely detecting a signal to identifying the underlying operator structure (dipole, Higgs-current, or four-fermion) by analyzing the geometry of the global likelihood and the correlation patterns among operators.

2. Methodology

The authors develop a unified, global statistical framework that integrates low-energy constraints with high-energy collider data. Key methodological components include:

• EFT Parameterization:

  • Uses a broken-phase EFT Lagrangian including dipole ($O_D$), Higgs-current ($O_{H\ell}, O_{He}$), and four-fermion ($O_{4f}$) operators.
  • Defines dimensionless reduced coefficients $c_\alpha = C_\alpha (1 \text{ TeV}/\Lambda)^2$.
  • Maps SMEFT operators at the electroweak scale to these reduced coefficients.

• Global Statistical Framework:

  • Likelihood Construction: A profile-likelihood ratio approach combining Poisson statistics for signal regions and Gaussian/log-normal priors for nuisance parameters (luminosity, background normalization, shape, theory uncertainties).
  • Differential Observables: Instead of inclusive rates, the analysis utilizes binned templates based on differential kinematic variables ($m_{\ell\ell}$, $p_T$, angular observables, beam polarization tags).
  • Machine Learning: Gradient-Boosted Decision Trees (GBDT) are trained to classify LFV signals against backgrounds. The classifier score is binned and included directly in the likelihood to maximize shape discrimination.

• Simulation and Validation:

  • Hybrid Strategy: Hadron and muon collider channels use event-level generation ($MadGraph5 \to Pythia8 \to Delphes$) with detector simulation. Lepton collider channels use CDR-calibrated parametric response models.
  • Validation: A specific channel ($pp \to e\mu$) is validated with full showering to quantify migration effects and tail depletion, which are propagated as correlated nuisance parameters to other channels.

• Advanced Theoretical Inputs:

  • RG Evolution: One-loop Renormalization Group (RG) running is applied to evolve Wilson coefficients from the UV matching scale ($\Lambda$) to the measurement scale ($\mu = m_Z$). This accounts for operator mixing (e.g., dipole mixing into four-fermion operators).
  • Polarization: Beam polarization asymmetries are used to disentangle chiral structures ($c_{H\ell}$ vs. $c_{He}$).
  • Dalitz Analysis: Differential $\mu \to 3e$ information is included via Dalitz plot templates to distinguish operator shapes in the low-energy sector.

• Metrics for Identification:

  • Fisher/Hessian Matrix: Used to quantify the local curvature of the likelihood. The correlation matrix $|\rho_{\alpha\beta}|$ identifies degeneracy directions.
  • Bayes Factors: Used for global model discrimination (e.g., Leptoquark vs. Heavy Neutral Lepton hypotheses) via marginal likelihood integration.
  • Information Gain: An optimization metric ($G$) based on the determinant of the Fisher matrix is used to optimize collider run plans (luminosity vs. polarization).

3. Key Contributions

1. Shift from Exclusion to Identification: The paper establishes a formalism where the "figure of merit" is the ability to break degeneracies in the Wilson coefficient space, rather than just maximizing the exclusion limit on $\Lambda$.
2. RG-Aware Global Fit: It is one of the first global analyses to consistently include one-loop RG evolution and operator mixing between UV matching and measurement scales, finding 10–30% shifts in operator correlations at multi-TeV scales compared to tree-level approximations.
3. Unified Low-Energy/Collider Framework: It successfully merges low-energy constraints (muon conversion, rare decays) with high-energy tails and resonance channels into a single statistical inference chain, demonstrating geometric complementarity.
4. Detector-Informed Benchmarks: Provides a reproducible code chain with event-level Delphes samples for hadron/muon colliders and validated parametric models for lepton colliders, including specific nuisance treatments for showering effects.
5. Run-Plan Optimization: Demonstrates that a balanced strategy (combining high-energy tails with polarization-resolved intermediate-energy data) yields a ~40% increase in information volume compared to symmetric or single-mode optimization strategies.

4. Key Results

• Complementarity: Low-energy experiments strongly constrain the dipole direction ($c_D$), while collider high-mass tails constrain the four-fermion direction ($c_{4f}$). Their combination contracts the allowed parameter space significantly more than the sum of individual parts (geometric rotation of constraints).
• Operator Separation:
  • Polarization asymmetries at ILC/CLIC effectively separate $c_{H\ell}$ and $c_{He}$, reducing the allowed region in the $(c_{H\ell}, c_{He})$ plane.
  • Dalitz-level $\mu \to 3e$ templates provide distinct shape discrimination between dipole, current, and contact interactions, which inclusive branching ratios cannot achieve.
• Model Discrimination:
  • The global combination achieves a Bayes factor of $\ln B_{LQ/HNL} \approx 5.41$, moving from "weak/moderate" evidence (single frontier) to "strong" evidence for model discrimination.
  • RG evolution is shown to rotate the preferred directions in coefficient space by amounts comparable to projected uncertainties, necessitating its inclusion for accurate UV interpretation.
• Systematic Robustness: The analysis shows that global combinations are more robust to systematic uncertainties (degrading separability by a factor of ~2 for hadron-only fits vs. much less for global fits) due to orthogonal information from different collider types.
• EFT Validity: Stress tests with validity clipping (suppressing events where $q^2 \sim \Lambda^2$) show that while absolute reach decreases, the qualitative complementarity pattern remains intact.

5. Significance

This work provides a blueprint for future collider physics programs focused on precision BSM discovery. It moves beyond simple "reach" projections to a sophisticated tomographic approach that:

• Quantifies how different experiments complement each other to solve the inverse problem of BSM physics.
• Provides a reproducible, detector-informed toolchain for experimental collaborations to optimize run plans (balancing luminosity and polarization).
• Offers a direct pathway from EFT fits to UV model rejection/ranking using Bayesian methods.
• Highlights that operator identification is the critical bottleneck for understanding new physics, requiring a global, multi-frontier approach that leverages differential observables, polarization, and RG evolution.

The paper concludes that future collider strategies must prioritize identifiability (breaking degeneracies) over raw significance, as this yields the most robust constraints on the nature of new physics.