⚛️ phenomenology

Interference effects in new physics searches

This review examines the critical impact of interference effects on new physics searches, particularly within extended scalar models, highlighting how their frequent neglect in current theoretical and experimental frameworks can lead to inaccurate descriptions of kinematical distributions in UV-complete models.

Original authors: Tania Robens

Published 2026-07-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Tania Robens

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Invisible Dance of Particles

Imagine you are at a massive, chaotic concert where thousands of people are dancing. If you want to find a specific group of dancers wearing bright red shirts, you might think the easiest way is to just count everyone in red. But what if the red-shirted dancers are constantly bumping into, pushing, or pulling on the people in blue shirts? What if, because of this interaction, the red shirts sometimes look like they are glowing brighter, or sometimes they seem to vanish entirely, blending into the crowd? In the world of particle physics, this is exactly what happens when scientists try to find new, heavy particles. They aren't just looking for a single "red shirt" appearing out of nowhere; they are watching a complex dance where the new particle (the signal) interacts with the background noise of known particles.

This paper dives into a specific corner of that dance floor: the search for "new physics" beyond our current best theory, the Standard Model. The key concept here is interference. In quantum mechanics, particles act like waves. When two waves meet, they can either add up to make a bigger wave (constructive interference) or cancel each other out to make nothing (destructive interference). For decades, scientists often tried to simplify their calculations by assuming these waves didn't really mix, treating the new particle as if it appeared, did its job, and disappeared without bothering the background. This paper argues that this simplification is a dangerous mistake. Just like ignoring the crowd's reaction would give you a wrong count of the red-shirted dancers, ignoring interference gives physicists a wrong picture of what the universe is actually doing.

The Paper's Big Reveal: The Ghost in the Machine

In this review, Tania Robens from the Rudjer Boskovic Institute pulls back the curtain on a common habit in physics research: the tendency to ignore the messy, wave-like interactions between new particles and the known ones. The paper suggests that many current searches for heavy, new particles are missing a crucial piece of the puzzle because they treat these new particles as if they are isolated islands, completely separate from the "background" noise of the Standard Model.

Robens explains that when a new, heavy particle is created in a collider (like the Large Hadron Collider), it doesn't just pop into existence and decay. It exists in a quantum superposition with other processes that look exactly the same. Think of it like two radio stations broadcasting on the same frequency. If you only tune into one, you might think you hear a clear song. But if you realize the other station is broadcasting a different song at the same time, the result is static, a weird echo, or a completely different melody. The paper shows that when you add the "static" (the interference) to the "song" (the new particle), the shape of the data changes dramatically.

The author walks us through several scenarios where this matters. For instance, when looking for a heavy particle that decays into pairs of W or Z bosons (which are like heavy cousins of light), the data doesn't look like a neat, smooth hill (a "Breit-Wigner" shape) that physicists expect. Instead, the interference creates dips, bumps, and weird distortions. In some cases, the interference is so strong that it cancels out the signal entirely in certain areas, making a heavy particle look like it doesn't exist at all. In other cases, it makes the signal look much stronger or shifts its position.

The paper highlights that this isn't just a tiny detail; it's a game-changer. The author points out that if you train a computer (like a neural network) to spot new physics using only the "clean" signal without the interference, the computer will be confused when it sees the real data. It might miss a discovery or, worse, think it found something when it's just a trick of the waves. The review covers various models, from simple extensions with one extra particle to more complex ones with two extra particles, and in almost every case, the interference effects are significant.

Robens also discusses how this plays out in different "final states," or what the particles turn into after they decay. Whether they turn into pairs of bosons, pairs of Higgs bosons, or even pairs of top quarks, the interference effects are there, often distorting the mass distribution in ways that a simple "signal-only" model cannot predict. In one extreme example involving two nearly identical heavy particles, the interference is so destructive that the signal completely disappears, leaving a flat line where a peak should be. This is a "ghost" effect where the new physics is there, but the waves cancel it out so perfectly that it looks like nothing happened.

The paper concludes with a strong recommendation: stop ignoring the dance. The tools to calculate these interference effects are already available, and experimental teams should use them. The author suggests that by including these "messy" interactions, physicists can avoid false negatives (missing a discovery) and false positives (thinking they found something that isn't there). It's a call to move from a simplified, idealized view of the universe to a more realistic, complex, and accurate one. The message is clear: if you want to find the new physics hiding in the noise, you have to understand how the noise and the signal are dancing together.

Technical Summary: Interference effects in new physics searches

Problem Statement
In searches for new physics (NP) involving additional scalar bosons that decay resonantly into Standard Model (SM) or new physics final states, a common theoretical and experimental assumption is that these resonances can be modeled separately and added incoherently to the SM background. This approach relies on the Narrow Width Approximation (NWA) and a factorized description of the process ( $a b \to S \to c d$ ). However, this review highlights that such descriptions neglect quantum mechanical interference effects between the signal (resonant) amplitude and background (non-resonant) amplitudes, as well as interference between multiple resonances. Neglecting these terms can lead to incorrect descriptions of kinematic distributions, particularly away from the resonance peak, potentially compromising the validity of signal modeling, cut optimization, and the interpretation of experimental limits within UV-complete models.

Methodology
The paper provides a review of existing literature and theoretical frameworks to demonstrate the necessity of including interference terms.

Theoretical Framework: The author revisits the foundations of Quantum Field Theory (QFT), contrasting the factorized NWA approach with the full quantum mechanical description where the total matrix element is the sum of the resonant signal ( $M_S$ ) and non-resonant/background contributions ( $M_{rest}$ ). The total cross-section involves the interference term $2 \text{Re}[M_S M^*_{rest}]$ , which is often omitted.
Model Scope: The review focuses on scalar extensions of the SM, specifically:
- Real Singlet Extension: An SM singlet scalar mixing with the Higgs doublet.
- Two Higgs Doublet Models (2HDM): Including various Yukawa types and symmetry classes.
Case Studies: The paper analyzes specific final states using results from previous studies (e.g., OpenLoops+Sherpa frameworks, various NLO calculations) to quantify interference effects. The analysis covers:
- Diboson final states ($WW, ZZ$).
- Di-Higgs final states ($hh$).
- Triple Higgs final states ($hhh$).
- Difermion final states ( $t\bar{t}$ ).
Observables: The impact of interference is assessed on invariant mass distributions ( $M_{VV}, M_{hh}, M_{t\bar{t}}$ ), transverse momenta ( $p_T$ ), and angular observables. The review also considers the effects of detector smearing and the use of advanced analysis tools like neural networks.

Key Contributions and Results
The review synthesizes evidence showing that interference effects are non-negligible and can drastically alter phenomenological predictions:

Diboson Channels: In singlet and 2HDM scenarios, the invariant mass distributions for $VV$ final states deviate significantly from the pure Breit-Wigner shape when interference with the continuum and SM-like scalars is included. These deviations are most pronounced away from the resonance mass ( $m_{VV} \neq m_H$ ).
Di-Higgs Channels: For $pp \to hh$ , interference terms modify the total rate and the shape of the invariant mass distribution. The review notes that detector smearing can cause SM-like contributions to mimic a second mass peak. Furthermore, ignoring interference can lead to significant differences in projected discovery reach and exclusion limits at the HL-LHC, depending on model parameters like $\tan \beta$ .
Triple Higgs Channels: In resonance-enhanced triple Higgs production ( $pp \to h_3 \to h_2 h_1 \to h_1 h_1 h_1$ ), interference is crucial for modeling transverse momentum distributions, where signal-only models fail to describe the intermediate kinematic regions.
Difermion Channels ( $t\bar{t}$ ): In 2HDMs with nearly mass-degenerate scalars, interference can lead to destructive effects. In extreme cases, the interference between two resonances and the background can result in a flat plateau in the background-subtracted invariant mass distribution, effectively canceling the resonance signal despite the presence of two s-channel states.
Quantification: The paper cites specific studies (e.g., CMS Run 2 combinations) that define ratios to quantify the magnitude of interference, showing that regions excluded under signal-only assumptions may require re-evaluation using full simulations.

Significance and Claims
The paper asserts that interference effects are a fundamental consequence of quantum mechanics and quantum field theory that cannot be a priori neglected, even in scenarios with narrow widths.

Correct Modeling: The author argues that leading-order tools capable of including these interference terms are readily available and should be utilized by experimental collaborations for signal modeling.
Impact on Analysis: The review emphasizes that neglecting interference leads to incorrect kinematic distributions, which can mislead neural network training and cut optimization strategies.
Modest Scope: The paper positions itself as a review of the current status and existing studies rather than a proposal of new experimental searches. It acknowledges that while leading-order tools are available, the correct treatment of higher-order corrections in production mechanisms remains an open topic, though proposals exist in the literature.
Call to Action: The primary conclusion is a recommendation for experimental collaborations to include interference effects in their studies, either through event reweighting or appropriate full simulations, to ensure a correct description of UV-complete physics scenarios.

The Invisible Dance of Particles

The Paper's Big Reveal: The Ghost in the Machine

More like this