Entropy and mean multiplicity from dipole models in the high energy limit

arXiv:2604.17418 · hep-ph · Submitted 2026-04-19 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: High Energy Physics: Phenomenology. The day in particle phenomenology: models, signatures and constraints.

Kai: We're looking at the paper: "Entropy and mean multiplicity from dipole models in the high energy limit".

Mira: Here's the gist of the paper. The paper investigates entropy and mean multiplicity from dipole models in the high energy limit to address ambiguities in experimental measurements by proposing a universal observable, S(ln⟨n⟩), which is compared against data from proton-proton collisions and different model calculations Universal Observable Definition The authors propose the entropy as the function of the logarithm of the average multiplicity, S(ln⟨n⟩), as a universal observable.

Kai: This function is defined by calculating it from experimental data using The definition is given by ⟨n⟩=X n P(n) and Sh=-X n P(n) ln(P(n)).

Mira: They obtain this universal S(ln⟨n⟩) from charged particle multiplicity distribution measured in p+p collisions and compare them to model calculations such as the 1D Mueller dipole model.

Kai: Dipole Model Descriptions The utilized dipole models are the 1D Mueller model

18, 19, 20: and its generalization

21, 16: which describe parton cascades that develops in the rapidity dimension.

Title and authors: Kai: So, this paper, "Entropy and mean multiplicity from dipole models in the high energy limit," it’s looking at how we measure things in proton-proton collisions. It tries to clean up some confusion around different ways people define the pseudorapidity ranges for those measurements.

Mira: Right, so they are proposing something called a universal observable, S(ln⟨n⟩), which is basically the entropy as a function of the logarithm of the average particle number. It’s supposed to be a standard way to compare different experimental results and model calculations.

Lev: That sounds like trying to find one single yardstick when you have so many different rulers being used by various experiments and theorists. I wonder how reliable this universal observable actually is when you're dealing with real, messy data from colliders.

Kai: Exactly, because they define it directly from the measured charged particle multiplicity distributions using the average number of particles, which they call n times probability n P of finding n particles at a certain spot.

Mira: And then they compare this universal S(ln⟨n⟩) against calculations from models like the 1D Mueller dipole model and its generalization <ref:2604.17418#pg1,the 1D Mueller dipole model>. The point is that the generalized model seems much better at describing the data than the original one, especially when looking at higher multiplicities.

Lev: If we're talking about fitting parameters based on fits in section three, does that mean they are using these models to predict things like how many partons you should expect to see in a collision?

Kai: Right, that’s what they do. The generalized model predicts a higher average number of partons and a higher entropy in the rapidity region being studied. They found the solution for this generalized model looks like a negative binomial distribution, which is pretty interesting because it's more dispersed than the original Mueller model.

Mira: That dispersion is key; it means the generalized model aligns much better with what’s actually measured in proton-proton collisions across different energy levels and ranges. They found that the distributions from the generalized model closely resemble those measured data in p-p collisions, unlike the original 1D Mueller distribution, which they said shows deviation at lower multiplicities <ref:2604.17418#pg1>.

Lev: So, if I'm building a simulation for real hardware based on this generalized model, does that mean my error bars will be smaller when I compare it to the experimental data they used?

Kai: Not necessarily smaller errors overall, but the fit itself is much better because the underlying physics description is more accurate for what you are measuring. They also noted that this model can describe data spanning two orders of magnitude in energy and covers a large pseudorapidity interval.

Mira: That wide range is important because it means this approach isn't just a tiny piece of the puzzle; it’s robust across different collision conditions. This work, "Entropy and mean multiplicity from dipole models in the high energy limit," aims to resolve those ambiguities in experimental measurements by giving us this standardized observable.

Lev: It sounds like a solid way to standardize the comparison, which is what we need when we're trying to build tools that can predict particle production with confidence. But what about the limitations of using this specific entropy function?

Kai: They explicitly state that the generalized model has an additional parameter that allows for a more accurate description of multiplicity, which is how they move beyond the original model. However, they also mention that extending this to include recombination effects would be valuable if we want to study saturation properly.

Mira: That's where it gets interesting. They suggest that if you extend this generalized model to take into account recombination effects, it could significantly improve the description of saturation behavior when we look at more complex colliding environments than standard deep inelastic scattering processes.

Lev: I agree, incorporating those recombination effects would definitely make the simulation more realistic for what we're trying to build with error correction research. It moves us from a simpler cascade picture to something closer to reality in terms of how particles interact and evolve.

Kai: So, this paper gives us a better tool—the universal S(ln⟨n⟩)—to compare what we measure across different experiments and models, and it points the way toward incorporating recombination for better saturation physics down the line. We can look forward to seeing if those extensions work.

Mira: It really seems like a foundational step in using this observable consistently, which should help us reduce those experimental ambiguities that have been bothering people when they compare measurements from ALICE or ATLAS with different theoretical frameworks.

Lev: For my perspective, the immediate value is in providing a better baseline model for multiplicity distributions that we can actually test against existing data sets right now. It’s a stepping stone toward more complex physics.

Kai: Well, that’s the picture on this paper, "Entropy and mean multiplicity from dipole models in the high energy limit." It solidifies a way to look at particle production through entropy and moves us closer to better understanding saturation in these high energy limits.

Mira: Yeah, it gives us a consistent language for talking about these things across different experiments and calculations. Good work by Kutak, Lóos, and the team on this paper.

The paper's summary: Kai: So we just looked at how they defined this universal entropy thing, S of n log n, and now we’re getting to what they actually found in their summary on page fifteen.

Mira: They're saying that this S of n log n is a standardized way to compare whatever experimental results you get from proton-proton collisions with the theoretical models you’ve been running.

Kai: Right, it boils down to them taking the measured particle count data and calculating this entropy value directly using those charged particle multiplicity distributions.

Mira: And they’re using that same tool to test things like the one-dimensional Mueller dipole model and its more complicated versions of it.

Kai: The main point is that they found their generalized model, which is a bit more complex than the original one, actually describes the experimental data much better than just the simpler Mueller model does.

Mira: They say this means their generalized version predicts a higher average number of partons and a higher entropy in the range of rapidities they’re studying.

Kai: It’s like saying that when you look at high energy collisions, this more detailed description gives you a better picture of how many particles are actually involved.

Mira: And they say the distributions from this generalized model look much more spread out, which is what matches the measured data in proton-proton collisions pretty well.

Kai: So what does that mean for us? It suggests that maybe our current models for particle production at high energy are missing some of those details.

Mira: It opens up a path where we can use this universal observable to compare things consistently across different experiments and different theoretical ideas, which should help clear up some of the confusion in the field.

Kai: If this model holds up across two orders of magnitude in energy and covers a wide range of pseudorapidity, that gives us confidence that it’s not just a fluke for one specific collision setup.

Mira: Exactly, they emphasize that this kind of study is really important if we want to push these models further, especially when we start thinking about saturation effects.

Kai: They hint that if the model gets extended to include recombination effects, then it could help us better describe what happens in those very complex colliding environments.

Mira: That’s the next big thing for this work; they aren't just stopping at describing multiplicity, they are setting up a framework for how we should look at saturation physics.

Kai: So, it’s a consistent way to measure entropy that points us toward where we need to go next in understanding these high energy collisions.

The paper's improvements: Kai: So we’re moving on to what the authors say they should do next, what their suggested improvements are for this whole entropy thing.

Mira: They’re suggesting that they need to take their current generalized model and extend it further by adding recombination effects.

Kai: Recombination effects sound like when particles bump into each other and stick together, which is something we have to account for in real high energy collisions.

Mira: Right, because they point out that if you want a really accurate picture of saturation, you can’t just ignore those interactions between the partons.

Kai: That means the next step isn't just fitting better numbers to the same data set; it’s changing the physics inside the model to match what we expect in a dense state.

Mira: They argue that this extension would allow them to predict saturation behavior in those more crowded, complex colliding environments we can’t easily recreate.

Kai: If they get that recombination part right, then this whole dipole model approach could actually become a better way to predict the particle yield at much higher energies than what we can do now.

Mira: It’s about moving from just describing the decay cascade to modeling how those particles actually interact and form a dense medium.

Kai: From my side, if we’re talking about experimental setup, that means our detectors need to be sensitive enough to see the subtle changes in particle interactions caused by recombination.

Mira: And for theory, it gives them a clear target—how the entropy should change when those recombination terms kick in.

Kai: So the implication is that this research isn't just about getting a better fit for today’s data; it’s laying out the blueprint for how to model future, denser collision states.

Mira: It’s about using this universal observable not just to describe what we see now, but to predict what we should see when the system gets really crowded.

Kai: And that leads right into the next question: how do we actually build the hardware or simulations that can test these kinds of complex interaction models?

Conclusion: Kai: So we’re wrapping up this discussion on "Entropy and mean multiplicity from dipole models in the high energy limit," summarizing what this paper actually accomplishes and where it leaves us next.

Mira: Basically, they established a consistent way to measure entropy using S of n log n, which lets us compare different experimental results with their dipole model calculations.

Kai: They proved that their generalized model provides a much better description of the particle data than the simpler version they started with.

Mira: It shows that when you use this more detailed framework, you get a prediction for average parton number and entropy that matches what we measure in proton-proton collisions across two orders of magnitude in energy.

Kai: It’s a solid tool for standardizing our approach to these measurements, which is something we really needed to get past some of the ambiguity in the field.

Mira: They conclude that this work is valuable because it sets up a robust baseline for studying saturation effects when they eventually extend the model to include recombination terms.

Kai: So, the main implication is that we have a better way to look at particle production data right now, even if we still need those recombination physics extensions later on.

Mira: They suggest that this paper gives us a consistent language for talking about these observables across different experiments and calculations, which should help reduce those experimental ambiguities.

Kai: It’s about getting our foundational measurements and theoretical predictions to speak the same standardized language before we try to tackle those super complex physics like saturation.

Lev: From my end, it means if we’re building error-correction hardware for real data analysis, having a better baseline model that is more dispersed than the original Mueller model helps us set up much more realistic simulation parameters.

Mira: Exactly, and that improved description of the distributions directly translates to better input for our theoretical calculations concerning medium effects.

Kai: So we have this solid foundation now to test those new recombination ideas when they come along, which is a big step forward for experimental planning.

Krzysztof Kutak, ł Sandor Lľ osľ

Institute of Nuclear Physics Polish Academy of Sciences · CPHT, CNRS, Ecole Polytechnique, Institut Polytechnique de Paris · Hungarian University of Agriculture and Life Sciences, Institute of Technology

hep-ph

Submitted: 2026-04-19

Updated: 2026-04-19

Comments: Proceedings of the 32nd Epiphany Conference, Krakow Poland that was organized in 12-16 of January, 2026

Journal ref: Acta Phys. Pol. B Proc. Suppl. 19, 5-A19 (2026)

DOI: 10.5506/APhysPolBSupp.19.5-A19

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 76/100

The gist: The paper investigates entropy and mean multiplicity from dipole models in the high energy limit to address ambiguities in experimental measurements by proposing a universal observable…

Key concepts

Universal Observable S(ln⟨n⟩)
This is a standardized way to measure entropy based on the logarithm of the average particle multiplicity (⟨n⟩). It is calculated using specific formulas from experimental data, allowing researchers to compare entropy values reliably across different experiments and theoretical models without issues related to varying pseudorapidity ranges.
Dipole Models
These are theoretical frameworks used to describe parton cascades that develop in the rapidity dimension during high-energy collisions. The paper focuses on the 1D Mueller model and its generalization, which aim to explain how partons interact and multiply as energy increases.
Generalized Dipole Model
This is an improved version of the dipole model that provides a significantly better description of experimental data than the original 1D Mueller model. It predicts a more dispersed distribution of particle multiplicities, closely resembling measured data in pp collisions across a wide range of energy and pseudorapidity.
Saturation Effects
This refers to phenomena expected in high-energy collisions where the density of partons becomes so high that they start interacting and recombining. Extending the generalized model to include these effects is crucial for accurately predicting saturation behavior in more complex collision environments.

Terminology

Summary

The paper investigates entropy and mean multiplicity from dipole models in the high energy limit to address ambiguities in experimental measurements by proposing a universal observable, S(ln⟨n⟩), which is compared against data from proton-proton collisions and different model calculations

Universal Observable Definition

The authors propose the entropy as the function of the logarithm of the average multiplicity, S(ln⟨n⟩), as a universal observable. This function is defined by calculating it from experimental data using

The definition is given by ⟨n⟩=X n P(n) and Sh=-X n P(n) ln(P(n)). They obtain this universal S(ln⟨n⟩) from charged particle multiplicity distribution measured in p+p collisions and compare them to model calculations such as the 1D Mueller dipole model. p. 1

Where the paper says it

This observable is straightforward to determine from the experimental data from their definition ⟨n⟩= X n nP(n) and Sh=− X n P(n) ln(P(n)). (1) We obtain the universal S(ln⟨n⟩) from charged particle multiplicity distribution measured in p+p collisions [8, 15, 10, 11, 12, 17, 13, 14] and compare them to model calculations such as the 1D Mueller dipole model [18, 19, 20] and one of its extension [21].

Dipole Model Descriptions

The utilized dipole models are the 1D Mueller model [18, 19, 20] and its generalization [21, 16] which describe parton cascades that develops in the rapidity dimension.

The 1D Mueller model is given by the cascade equation ∂yPn(y) = −αnPn(y) + (n − 1)αPn−1(y). p. 1

Where the paper says it

The generalized model has an additional parameter that allows for more accurate description of multiplicity. 1D Mueller model The 1D Mueller model that depends only on rapidity is given with the following cascade equation ∂yPn(y) = −αnPn(y) + (n − 1)αPn−1(y), (2) where the Pn(y) is the initial state parton multiplicity distribution and α represents the gluon emission kernel and characterizes the speed of the Balitsky–Fadin–Kuraev–Lipatov (BFKL) cascade evolution [22, 23].

The solution of this equation can be written in the form of Pn(y) = 1/C e−αy 1 − 1/C e−αyn−1. p. 1
Where the paper says it

1D Mueller model The 1D Mueller model that depends only on rapidity is given with the following cascade equation ∂yPn(y) = −αnPn(y) + (n − 1)αPn−1(y), (2) where the Pn(y) is the initial state parton multiplicity distribution and α represents the gluon emission kernel and characterizes the speed of the Balitsky–Fadin–Kuraev–Lipatov (BFKL) cascade evolution [22, 23]. The solution of the equation can be written in the form of Pn(y) = 1 C e −αy 1 − 1 C e −αyn−1 , (3) with the recently introduced C parameter [3].

The generalized model introduces a new parameter that allows for more accurate description of multiplicity. p. 1

Where the paper says it

The dipole cascade models The utilized dipole models are, the 1D Mueller model [18, 19, 20] and its generalization [21, 16] that describe parton cascades that develops, in the 1D case, in rapidity. The generalized model has an additional parameter that allows for more accurate description of multiplicity.

The solution has the form of the negative binomial distribution with the above defined p = 1/C e−αy probability. p. 2
Where the paper says it

Generalized Mueller model A new cascade equation that was introduced in Ref. [21] has a new parameter as ∂yPn(y) = −α(n + 2h)Pn(y) + α(n − 1 + 2h)Pn−1(y) (6) and its solution has the form of the negative binomial distribution with the above defined p = 1 C e −αy probability: Pn(y) = Γ(2h + n) n!

Model Comparison and Data Fit

The comparison shows that the generalized model can well describe the data while the Mueller model exhibits deviation from the data at lower multiplicities. The generalized model predicts higher average number of partons and higher entropy in the investigated rapidity range. p. 3

Where the paper says it

It can be seen in Fig. 2 that the generalized model predicts higher average number of partons and higher entropy in the investigated rapidity range.

The models are evaluated with parameters based on fits that will be discussed in Sec. p. 2

Where the paper says it

Comparison of the models We show the comparison of the two studied models in terms of logarithm of average multiplicity and entropy. The models are evaluated with parameters based on fits that will be discussed in Sec.

3. The solutions given in Eqs. p. 2
Where the paper says it

The models are evaluated with parameters based on fits that will be discussed in Sec. From the solutions given in

(3) and (7), the multiplicity distributions for the original dipole model features exponential decay, the distributions from the generalized model are more disperse that closely resemble the measured multiplicity data in pp. p. 3
Where the paper says it

The fixed values are based on earlier results [27]. (3) and (7), the multiplicity distributions for the original dipole model features exponential decay, the distributions from the generalized model are more disperse that closely resemble the measured multiplicity data in pp [8, 15, 10, 11, 12, 17, 13, 14] and ep [1].

Conclusion and Outlook

The generalized model can describe data spanning 2 orders of magnitude in energy and covers a large pseudorapidity interval. The authors emphasize that such studies could be valuable in the context of saturation if the model would be extended to take into account for recombination effects. p. 4

Where the paper says it

We found that the generalized model can describe the data while the Mueller model deviates from the data points at lower multiplicities. We would like to emphasis that such studies could be valuable in the context of saturation if the model would be extended to take into account for recombination effects.

Theoretical calculation could improve the description of pp data as more complex colliding environment than DIS processes. p. 4
Where the paper says it

We would like to emphasis that such studies could be valuable in the context of saturation if the model would be extended to take into account for recombination effects. Also, theoretical calculation could improve the description of pp data as more complex colliding environment than DIS processes.

Experimentally, the addressed methodological issue could be resolved and further, more detailed analyses could refine the present and future results. p. 4
Where the paper says it

Also, theoretical calculation could improve the description of pp data as more complex colliding environment than DIS processes. Experimentally, the addressed methodological issue could be resolved and further, more detailed analyses could refine the present and future results.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Use universal entropy S(ln⟨n⟩) as a standardized observable for comparing measured entropies across different experiments and model calculations, resolving ambiguities stemming from different definitions of pseudorapidity ranges. This allows AI systems to reliably compare experimental data from proton-proton collisions with theoretical predictions like the 1D Mueller model or its generalizations.

  2. Improve high energy limit modeling by utilizing the generalized dipole model, as it provides a significantly better description of the data than the 1D Mueller model. This enhanced modeling capability allows AI systems to predict average number of partons and entropy values more accurately across a wider range of experimental conditions.

  3. Develop predictive models for particle multiplicity distributions by incorporating the generalized Mueller model, which is described as being more disperse that closely resemble the measured multiplicity data in pp. This enables AI systems to better fit experimental charged particle multiplicity distributions from various collision energies and pseudorapidity ranges.

  4. Enhance saturation studies by extending the generalized model to include recombination effects, as suggested by the paper for future theoretical calculation improvements. This allows AI systems to predict saturation behavior in more complex colliding environments than standard deep inelastic scattering processes.

Abstract

The 1D Mueller dipole model, its high energy limit, and its generalization were investigated. To address the ambiguity stemming from different definitions of the pseudorapidity ranges in experimental measurements, we propose the entropy as the function of the logarithm of the average multiplicity, S(n, as a universal observable. From the solutions of the models, we calculate both the entropy and the average charged particle multiplicity and compare to data measured in proton-proton collisions. We obtained these quantities directly from the measured charged particle multiplicity distributions and determine the model parameters via fits. We find that the generalized dipole model provides a significantly better description of the data than the 1D Mueller model.

Sources

Related papers