Is the Standard Model Effective Field Theory Enough for Higgs Pair Production?

This paper evaluates the validity of the Standard Model Effective Field Theory (SMEFT) versus the Higgs Effective Field Theory (HEFT) for Higgs pair production by comparing their predictions against "Loryon"-inspired UV-complete models, demonstrating that HEFT can offer a more accurate description than SMEFT in specific parameter regimes and highlighting di-Higgs measurements as a probe for non-linear electroweak dynamics.

The Gist

The Big Picture: Two Maps for the Same Territory

Imagine the universe is a vast, mysterious landscape. Physicists have a "Standard Map" (the Standard Model) that describes how particles behave. But they suspect there are hidden mountains and valleys they haven't seen yet (New Physics).

To explore these unknown areas without getting lost, scientists use Effective Field Theories (EFTs). Think of an EFT as a zoomed-out map. It doesn't show every single pebble (every tiny particle detail); instead, it shows the general shape of the terrain and how things flow over it.

This paper asks a crucial question: Which zoomed-out map should we use to study "Higgs Pair Production"?

There are two main maps on the table:

1. SMEFT (The "Linear" Map): This map assumes the landscape is smooth and predictable. It works great if the hidden mountains are very far away and heavy. It treats the Higgs boson like a standard citizen of the particle world.
2. HEFT (The "Non-Linear" Map): This map is more flexible. It assumes the landscape might have weird curves, sharp turns, or hidden structures that the smooth map misses. It treats the Higgs boson as a unique, independent traveler.

The Goal: The authors want to know: When we try to create two Higgs bosons at once (Higgs Pair Production), does the smooth map (SMEFT) work, or do we need the flexible map (HEFT) to avoid getting lost?

---

The "Loryon" Analogy: Where does the weight come from?

To test these maps, the authors invented a concept called "Loryons."

Imagine you are lifting a heavy backpack.

• Scenario A (SMEFT): The backpack is heavy because it's filled with bricks you brought from home (an "explicit mass"). The weight doesn't change much based on where you are.
• Scenario B (HEFT/Loryon): The backpack is heavy because you are standing on a giant, invisible trampoline (Electroweak Symmetry Breaking). The heavier the trampoline pushes back, the heavier the backpack feels. If more than half the weight comes from the trampoline, the "smooth map" (SMEFT) starts to fail because it doesn't understand how the trampoline works.

The authors tested three specific "backpacks" (models) to see which map works best:

1. The Single Extra Stone (Scalar Singlet Model)

• The Setup: They added one extra invisible stone to the universe.
• The Test: They checked if the stone got its weight from the trampoline or from being a heavy rock.
• The Result: When the stone got most of its weight from the trampoline (high "Loryon" status), the HEFT map was much more accurate. The SMEFT map tried to force a square peg into a round hole, leading to errors. However, if the stone was just a heavy rock (low trampoline influence), the smooth SMEFT map worked fine.

2. The Twin Towers (Two-Higgs-Doublet Model)

• The Setup: Imagine the universe has two Higgs bosons instead of one, like a twin tower.
• The Test: They looked at how these towers interact with top quarks (heavy particles).
• The Result: Similar to the first case. When the interaction was complex and relied heavily on the "trampoline" mechanism, the HEFT map described the reality much better. The SMEFT map started to break down, predicting results that didn't match the "real" twin towers.

3. The Colored Paint (Colored Scalar Model)

• The Setup: They added a particle that interacts with the "strong force" (like paint interacting with a wall).
• The Test: They checked how this paint affected the creation of Higgs pairs.
• The Result: In this case, the effect was so subtle and small that both maps looked almost the same. The difference between the smooth map and the flexible map was smaller than the "static" or noise in our current telescopes. So, for this specific case, the simpler SMEFT map is "good enough" for now.

---

The "Flare" Function: The Dealbreaker

The paper mentions a specific mathematical feature called the "flare function."

• Analogy: Imagine you are driving a car.
  • In the SMEFT world, the steering wheel is perfectly linear. If you turn it 10 degrees, the car turns 10 degrees. If you turn it 20 degrees, it turns 20.
  • In the HEFT world, the steering wheel is "non-linear." At low speeds, it's linear. But if you turn it too far, the car might suddenly swerve or drift in a way the linear map couldn't predict.

The authors found that in the "Loryon" scenarios (where particles get mass from the Higgs mechanism), the steering wheel does swerve. The HEFT map captures this swerve; the SMEFT map assumes the car will just keep turning smoothly and eventually crashes into a wall (mathematically breaking down).

The Conclusion: Why This Matters

The authors conclude that we cannot blindly trust the simple map (SMEFT) for all future experiments.

• For some models: The simple map is fine.
• For "Loryon" models: The simple map gives the wrong answer. If we use it to analyze data from the Large Hadron Collider (LHC), we might miss new physics or misinterpret what we see.

The Takeaway:
As we build better telescopes to watch Higgs bosons collide, we need to be ready to switch from the "Smooth Map" (SMEFT) to the "Flexible Map" (HEFT). If we stick to the smooth map when the terrain is actually bumpy, we might think we've found a new particle when we've just made a mapping error.

In short: The paper is a warning label for physicists. It says, "Check your map! If the new particles are 'Loryons,' the standard map is too simple, and you need the more complex one to find the truth."

Technical Summary

1. Problem Statement

Following the discovery of the Higgs boson, the Standard Model (SM) remains consistent with experimental data, motivating the use of Effective Field Theories (EFTs) to parameterize potential new physics (NP) in a model-independent way. Two primary frameworks exist:

• SMEFT (Standard Model EFT): Assumes the Higgs field is part of an $SU(2)$ doublet. Operators are ordered by canonical dimension.
• HEFT (Higgs EFT): Treats the physical Higgs as a singlet and Goldstone bosons non-linearly. Operators are ordered by chiral dimension.

The Core Question: Is SMEFT sufficient to describe Higgs pair production ($hh$), or are there specific UV scenarios where HEFT is required for an accurate description?

• Context: While $hh$ production is often used to probe the trilinear Higgs coupling, it is also a sensitive probe of non-linear electroweak dynamics. In SMEFT, couplings of one Higgs to vector bosons/fermions are correlated with two-Higgs couplings at a given order. In HEFT, these are independent.
• Motivation: The authors investigate "Loryon" models—scenarios where new states acquire more than half their mass from Electroweak Symmetry Breaking (EWSB). In such cases, the SMEFT expansion (around $H^\dagger H = 0$) may fail to converge, necessitating HEFT.

2. Methodology

The authors compare the predictions of full UV-complete models against their low-energy EFT counterparts (SMEFT at dimension-6 and LO HEFT) for Higgs pair production in two channels: Gluon Fusion (ggF) and Vector Boson Fusion (VBF).

Models Studied

Three benchmark UV models were selected, inspired by the Loryon catalogue:

1. Scalar Singlet Extension: Adds a real scalar $\Phi$ mixing with the SM Higgs. Generates tree-level contributions to kinetic mixing ($C_{H,\square}$) and the Higgs potential ($C_H$).
2. Two-Higgs-Doublet Model (2HDM): Adds a second $SU(2)$ doublet. Generates tree-level contributions to top-Yukawa modifications ($C_{tH}$) and the Higgs potential.
3. Colored Scalar Extension: Adds a colored scalar $\omega_1$ (charge $1/3$). Generates loop-induced contributions to the Higgs-gluon coupling ($C_{HG}$).

Technical Approach

• Matching: The authors perform tree-level matching for the Singlet and 2HDM, and one-loop matching for the Colored Scalar. They derive Wilson coefficients for SMEFT and coupling modifiers ($c_{VV}, c_{VVV}, c_{ff}, c_{gg}$) for HEFT.
• Parameter Space Scans: They scan the parameter spaces of these models subject to theoretical constraints (vacuum stability, perturbativity, unitarity) and experimental constraints (Higgs coupling modifiers, $W$-mass, flavor physics).
• Comparison Metric: They calculate the ratio of the cross-section (or amplitude squared) predicted by the full UV model ($\sigma_{UV}$) against the EFT predictions ($\sigma_{SMEFT}$ and $\sigma_{HEFT}$).
  • Ratio $R = \sigma_{UV} / \sigma_{EFT}$. A value close to 1 indicates a good EFT description.
• Kinematics:
  • ggF: Calculated at Leading Order (LO) in QCD using the infinite top-mass limit (K-factors cancel in ratios).
  • VBF: Calculated via unpolarized amplitude squared for $VV \to hh$ at fixed center-of-mass energies.

3. Key Contributions

• Quantitative Criteria for EFT Validity: The paper establishes specific regimes (defined by mixing angles, mass hierarchies, and symmetry breaking parameters) where SMEFT breaks down and HEFT becomes necessary.
• Loryon Analysis: It explicitly tests the "Loryon" hypothesis, showing that when new states derive significant mass from EWSB (parameterized by a factor $f$), the SMEFT expansion radius is exceeded.
• Coupling Decorrelation: The study highlights how HEFT naturally decouples single-Higgs and double-Higgs couplings, whereas SMEFT enforces correlations that can lead to significant deviations in the UV limit.
• Field Redefinitions: The authors carefully handle field redefinitions (specifically removing derivative self-interactions) to ensure a fair comparison between the EFT bases.

4. Results

A. Scalar Singlet Model

• Mechanism: Mass generation via EWSB is controlled by the parameter $f = \frac{\lambda_3}{\lambda_3 + \lambda_{ex}}$. As $f \to 1$, the singlet mass comes entirely from EWSB.
• Findings:
  • Small $f$ / Small $v_S$: SMEFT and HEFT agree well.
  • Large $f$ / Large $v_S$: SMEFT fails. The ratio $\sigma_{UV}/\sigma_{SMEFT}$ deviates significantly from 1, and in some regions, $\sigma_{SMEFT}$ becomes negative (indicating a breakdown of the EFT expansion).
  • Conclusion: HEFT provides a significantly better description for large $f$. The discrepancy is driven by terms proportional to $(v_H/v_S)\theta^3$ which are suppressed in SMEFT but present in HEFT.

B. Two-Higgs-Doublet Model (2HDM)

• Mechanism: Deviations are driven by the mixing angle $\alpha - \beta$ (specifically $c_{\beta-\alpha}$) and the heavy mass scale $M^2$.
• Findings:
  • Type I vs. Type II: Type II models are more constrained by flavor physics and Higgs coupling measurements, limiting the parameter space. Type I allows larger deviations.
  • Agreement: For small $c_{\beta-\alpha}$, both EFTs work. However, as $c_{\beta-\alpha}$ increases (moving away from the alignment limit), SMEFT predictions diverge from the UV model, while HEFT remains accurate.
  • Interference: In regions with negative $c_{\beta-\alpha}$, the interference between the continuum and the heavy Higgs resonance is delicate. SMEFT truncation at $O(1/\Lambda^2)$ fails to capture this interference correctly, whereas HEFT handles the non-linear structure better.

C. Colored Scalar Model

• Mechanism: Loop-induced coupling to gluons.
• Findings:
  • The difference between SMEFT and HEFT is minimal because the mass of the colored scalar is dominated by an explicit mass term ($M_{ex}^2$), keeping the system in the SMEFT regime.
  • The cross-section deviation from the SM is small (below theoretical uncertainties), making it difficult to distinguish the EFTs experimentally in this specific scenario.

5. Significance and Conclusions

• HEFT Superiority in Specific Regimes: The paper demonstrates that for models where new states acquire mass primarily through EWSB (Loryons), HEFT is not just a generalization but a necessary tool for accurate phenomenology. SMEFT can yield unphysical results (negative cross-sections) or large errors in these regimes.
• Experimental Implications: Current LHC searches for Higgs pair production often assume SMEFT. If the underlying physics involves Loryons, these analyses might misinterpret the data or miss signals. Di-Higgs measurements are identified as a crucial probe for non-linear electroweak dynamics.
• Validity Limits: The study clarifies that the validity of SMEFT is not just about the energy scale $\sqrt{s}$ but also about the origin of mass for the new particles. If the mass is "soft" (from EWSB), the SMEFT expansion parameter is not small, and HEFT is required.
• Future Directions: The authors note that their study is limited to LO HEFT and dimension-6 SMEFT. Future work should include higher-order corrections and loop effects in the matching to fully quantify the convergence of these theories.

In summary, the paper provides a rigorous phenomenological case that HEFT is required for a correct description of Higgs pair production in specific, theoretically well-motivated UV scenarios, challenging the default assumption that SMEFT is sufficient for all BSM searches.