Two-loop gg to AA amplitudes near threshold and the double parton scattering singularity

arXiv:2610.19024 · hep-ph · Submitted 2026-10-13 · 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: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Two-loop gg to AA amplitudes near threshold and the double parton scattering singularity".

Mira: Here's the gist of the paper. The two-loop amplitudes for pseudoscalar pair production in gluon fusion, gg → AA, with each pseudoscalar emitted from its own effective gluon vertex are known in closed form, but large cancellations near threshold obscure their analytic structure<ref:2610.19024#pg6>.

Kai: Threshold Expansion and Analytic Structure The paper expands the IR-finite amplitudes in the pseudoscalar velocity βA through β6A at fixed scattering angle<ref:2610.19024#pg4>.

Mira: All inverse powers of βA and the odd powers through β5A cancel, and at leading power the angular dependence enters only through ln(2pT /mA)<ref:2610.19024#pg6>.

Kai: This logarithm is identified with the double parton scattering (DPS) singularity<ref:2610.19024#pg4>.

Title and authors: Kai: So we're looking at this paper right now from arXiv:two thousand six hundred ten point one nine zero two four titled "Two-loop gg to AA amplitudes near threshold and the double parton scattering singularity." It sounds like they're tackling a real headache in calculating how often gluons fuse to make two pseudoscalar particles, gg to AA.

Mira: Exactly. The main point here is that those two-loop amplitudes are known in closed form, but when you look near the threshold where the particles are barely coming out, things get really messy because there are these large cancellations that hide what's actually happening analytically.

Lev: From an error correction standpoint, I’m curious about how they handle that mess. If the structure is obscured, it makes it hard to build reliable predictions for real hardware runs where we need high fidelity.

Kai: Well, what they do is expand those infrared finite amplitudes in terms of the pseudoscalar velocity beta A up to beta six A at a fixed scattering angle. They say all inverse powers of that velocity and the odd powers through beta five A just cancel out when you look at the leading power behavior.

Mira: That cancellation leaves only one thing visible at that leading power: an angular dependence that comes through the logarithm of two times the transverse momentum divided by the mass of a pseudoscalar, or ln(2pT over mA) <ref:2610.19024#pg1>.

Lev: That logarithm is what they identify as something called the double parton scattering singularity. I need to know if this structure is real enough to impact what we actually measure in a collider environment.

Kai: They show that at two loops, the equal helicity amplitude grows as ln three times the transverse momentum, and the opposite helicity amplitude grows as ln two times the transverse momentum.

Mira: That growth rate is significant because it tells us how sensitive our cross section predictions are to this specific kinematic region. It's not just a constant correction; it scales with that logarithm.

Lev: But what about running this on real hardware? If we have these logarithmic terms dominating, does the overhead for error correction become intractable quickly?

Kai: The paper is digging into the leading power terms and rewriting them in terms of some quantity called L, which is defined as ln beta A plus one half times the logarithm of one minus c squared plus ln two plus i pi over two.

Mira: That L term essentially becomes the logarithm of the transverse momentum up to a constant phase and some power corrections, which they link directly back to that double parton scattering singularity.

Title and authors: Lev: So, this means they are trying to map this theoretical singularity onto a physical observable that we can actually measure with our detectors.

Kai: The finite virtual corrections are quantified by dimensionless coefficients called V one and V two relative to the tree contribution. They find that the interference between the two-loop contribution and the tree amplitude gives a quadratic logarithm in its interference.

Mira: And from that, they get a specific relation for V two, showing it is proportional to negative seven over pi squared times N squared times c times ln two of twice the transverse momentum divided by mA.

Lev: That coefficient V two is what we need to worry about for precision; if it's large, it means our tree-level calculation isn't good enough near threshold.

Kai: They also show that the leading power expression for V one involves M of zero times beta A to the power of zero times X three k equals zero Abik L k.

Mira: That connects back to that logarithm we talked about, L, confirming that this term is intrinsically tied to the transverse momentum scale in a specific way.

Lev: That's interesting because it suggests that this scaling behavior isn't just an artifact of the calculation; it’s built into how these amplitudes behave at leading order.

Kai: The paper shows they have checked their results by expanding up to beta six A, and they reproduce the two complete remainders with maximum relative errors of three point two three times ten to the negative six for beta A less than or equal to zero point two.

Mira: That accuracy level is quite high for a series expansion, especially when you consider that these coefficients are even in cosine theta, which satisfies Bose symmetry order by order.

Lev: High accuracy in the numerical evaluation is good, but does that mean we can trust these results when we apply them to more complex scenarios or higher orders?

Kai: The series files they generated are actually a factor of five thousand sixty-one smaller than the ones from reference nine, which shows how efficient this method is at capturing the structure.

Mira: So, when you look at the forward limit, where you take c approaching one and the transverse momentum approaches zero, they probe the DPS singularity again. They found that in this limit, the one-loop DPS logarithm drops out of the tree interference because of their relative phase.

Title and authors: Lev: That's a key point for me; it means we can simplify our thinking about how these singularities manifest when particles move along the beam line instead of scattering off each other.

Kai: At two loops, they find that the J z equals zero and the absolute value of J z equals two remainders develop leading logarithms called L three and L two, respectively.

Mira: And they found that the equal helicity leading logarithmic ratio agrees with what we see in planar N four super Yang minus m theory. That comparison gives a strong validation point for their structure.

Lev: So the structure is consistent across different theoretical frameworks, which is reassuring when you're trying to build something that has to work on real hardware.

Kai: The forward limit calculation they do next shows that the same coefficient relation persists even away from threshold. The two-loop leading logarithms of M one in the forward limit are therefore the one-loop DPS logarithm, with its helicity weight FA of x multiplied by the same factor at every velocity.

Mira: That implies this logarithmic structure isn't just a threshold effect; it’s a more general property that dictates how these amplitudes behave across different kinematic regions.

Lev: If that's true, then our error correction models have to account for this universal scaling behavior instead of treating it as an isolated boundary condition.

Kai: The paper says this series provides the two loop Type two input for the hard function for the pair invariant mass distribution near mAA equals two times mA at next-to-next-to leading order at N three loop in a two h d m.

Mira: That means they are using this calculation to feed into higher order calculations, like n three loop in a two h d m, which is important for precision tests of the standard model.

Lev: It sounds like this work is setting up the necessary input to make those higher-order predictions actually feasible on real machines.

Kai: The paper recommends this threshold series as being good for zero point one, meaning it should be reliable right at the threshold region where we are focusing.

Mira: So, to wrap up, the paper "Two-loop gg to AA amplitudes near threshold and the double parton scattering singularity" shows that we can systematically expand these messy two-loop calculations near a kinematic edge.

Lev: The implication is that by expanding in terms of velocity beta A and isolating that specific logarithm, we explicitly identify the double parton scattering singularity structure.

Title and authors: Kai: And they show this structure dictates how the amplitude grows with transverse momentum at two loops, with the equal helicity amplitude growing as ln three times pT.

Mira: This helps us understand that the sensitivity to transverse momentum isn't just a simple power law; it has a logarithmic component tied to the DPS singularity.

Lev: For someone building error correction systems, this means we have a concrete way to quantify how sensitive our measurements will be near those kinematic boundaries.

Kai: We should also remember that the leading power terms are related to L, which is essentially the logarithm of the transverse momentum up to some constant phase and corrections.

Mira: And they've shown that this structure is consistent when we look at the forward limit, which means we can use this finding to guide our understanding away from just the threshold.

Lev: It’s about connecting a specific mathematical singularity to something physically measurable in our experiments.

Kai: So, if you're looking at precision calculations for Higgs production or other gluon fusion processes, this paper provides a concrete piece of input that addresses those two-loop challenges near threshold.

Mira: We need to be careful not just to calculate the main amplitude, but also to understand these finite virtual corrections quantified by V one and V two, because they are what introduce the quadratic logarithm in the interference.

Lev: If we can control those coefficients, we might have a better handle on the uncertainty in our predictions for real experimental data.

Kai: The whole point of this paper is to tame the analytic structure near threshold so that it becomes manageable enough to use as input for higher-order calculations in models like the two h d m.

Mira: So, we've seen how expanding in velocity helps us isolate the double parton scattering singularity, and how that structure dictates the growth of both helicity amplitudes at two loops.

Lev: It gives us a systematic way to handle these complex terms when we move from theory to what we would actually try to run on a real quantum error correction system.

Kai: The paper "Two-loop gg to AA amplitudes near threshold and the double parton scattering singularity" is all about taking those closed form results and making them useful for precision physics calculations.

Mira: We are leaving this one with the idea that the double parton scattering singularity isn't just some abstract mathematical quirk; it has a specific, predictable scaling behavior in terms of transverse momentum.

Lev: It’s about connecting that abstract singularity to concrete numerical scaling and error estimates for our hardware.

The paper's summary: Kai: So, this paper is about taking these two-loop calculations for when gluons fuse to make two pseudoscalar particles, gg to AA, right at the edge where they are barely coming out together and figuring out what's actually going on.

Mira: Yeah, it boils down to how those complex parts cancel out near threshold and what that cancellation reveals about a physical effect called the double parton scattering singularity.

Lev: I mean, it’s interesting because these amplitudes are notoriously messy when you get two loops in, and they hide the actual physics in those large cancellations.

Kai: Exactly. The authors expand the calculation using this velocity beta A up to beta six A at a fixed angle, and they show that all the inverse powers of that velocity and those odd powers just disappear at the leading power level.

Mira: And what remains at that leading power is this logarithm, ln of two times the transverse momentum divided by the mass of one pseudoscalar.

Lev: That logarithm is what they identify as a double parton scattering singularity, which is pretty significant because it’s a specific structure in how particles interact.

Kai: At two loops, they show that the equal helicity amplitude grows with ln three times the transverse momentum, while the opposite helicity amplitude grows with ln two times the transverse momentum.

Mira: That difference in growth rates is important because it tells us how sensitive our predictions are to this specific kinematic region when we look at different particle spins.

Lev: From an error correction standpoint, having these logarithmic terms scaling like that means we can't just treat them as small corrections; they have a definite power law dependence on the momentum scale.

Kai: They connect this back to some quantity L, which is basically the logarithm of the transverse momentum up to a constant phase and some power corrections.

Mira: The authors show that this structure isn't just a threshold artifact; it persists even when you look at the forward limit, where things are moving along the beam line instead of scattering off each other.

Lev: That’s interesting because if that scaling behavior holds across different kinematic regions, we can use it to guide our error correction models more generally.

Kai: They also show that these leading logarithmic terms at two loops match what we see in planar n four super Yang minus m theory, which gives them a lot of confidence in their math.

Mira: It’s pretty solid validation when you compare their results to established theories like that, but we still have to be careful about those finite virtual corrections quantified by V one and V two.

Lev: Those coefficients are what we need to worry about for real hardware; if they're large, it means the tree level calculation isn't good enough near threshold.

Kai: The paper gives us the two loop Type two input for calculations at next to next to leading order at n three loop in a two h d m model.

Mira: So, we’ve seen how expanding in velocity helps us isolate that double parton scattering singularity, and how that structure dictates the growth of both helicity amplitudes at two loops.

Lev: It gives us a systematic way to handle these complex terms when we move from theory to what we would actually try to run on real quantum error correction systems.

Kai: We're leaving this with the idea that this specific logarithmic scaling behavior is a universal property that dictates how these amplitudes behave across different kinematic regions.

Mira: We need to be careful not just to calculate the main amplitude, but also to understand those finite virtual corrections quantified by V one and V two because they introduce that quadratic logarithm in the interference.

Lev: If we can control those coefficients, we might have a better handle on the uncertainty in our predictions for real experimental data.

Kai: The whole point of this paper is to tame that analytic structure near threshold so it becomes manageable enough to use as input for higher order calculations in models like the two h d m.

Mira: So, we've seen how expanding in velocity helps us isolate the double parton scattering singularity, and how that structure dictates the growth of both helicity amplitudes at two loops.

Lev: It gives us a systematic way to handle these complex terms when we move from theory to what we would actually try to run on real quantum error correction systems.

The paper's improvements: Tom: So, we're looking at how this paper improves on the original two loop calculation for those gg to AA amplitudes near threshold, right?

Kai: The main thing they suggest is expanding that velocity series further, all the way up to beta six A. They say this allows them to reproduce the full amplitudes with very small errors when the velocity is less than zero point two.

Mira: That’s a big deal because it means they can get a more complete picture of what's happening right at that threshold region, rather than just getting an approximation early on.

Lev: From my side, that kind of high-order accuracy in the series expansion is valuable for checking if our error correction models are actually robust enough to handle the required precision.

Kai: They also show they have a much smaller series file compared to some other papers, which makes their method quite efficient for numerical evaluation.

Mira: And they refine how they identify that double parton scattering singularity by showing it appears explicitly when you expand those infrared finite amplitudes in terms of velocity beta A.

Lev: That’s the part that really matters for us; if we can explicitly isolate the singularity, we know exactly what structure we need to model and correct for on real hardware.

Kai: They also perform some forward limit tests on those logarithmic coefficients, confirming that the angular dependence at leading power only comes through that transverse momentum logarithm.

Mira: That confirms our earlier point; it means this scaling behavior is general, not just a trick specific to the threshold region.

Lev: But what about applying these results outside of threshold? Does this improved expansion still hold when we move away from zero?

Kai: They show that the same coefficient relations persist even away from threshold, which suggests that the physics described by those leading logarithms is more universal than just a boundary condition.

Mira: That’s a strong result; it means we can use this framework to predict how these amplitudes behave in other kinematic regions too.

Lev: If that holds up, it helps us build error correction protocols that account for this scaling behavior across different energy scales, not just near the threshold.

Kai: Ultimately, they use this improved series as the Type two input for hard functions in higher order calculations at next to next to leading order at n three loop in a two h d m model.

Mira: So, by improving the method and expanding it further, they’re giving us a better tool to calculate those high-precision inputs needed for testing models like the two h d m.

Lev: This is good because it gives us concrete input that we can use to see how much precision we actually gain on our hardware simulations.

Kai: So, the improvements are about getting higher accuracy at threshold and showing that the underlying physics structure is consistent across different kinematic regions.

Mira: We're leaving this with the idea that this refined series expansion gives us a more complete and reliable way to handle these complex two loop calculations for our models.

Conclusion: Kai: So, to wrap up, this paper on "Two-loop gg to AA amplitudes near threshold and the double parton scattering singularity" shows how we can systematically expand those messy two loop calculations using a velocity series.

Mira: They found that by expanding in velocity beta A, we can explicitly isolate that double parton scattering singularity structure which dictates how the amplitude grows with transverse momentum at two loops.

Lev: It’s important for us because it gives us a systematic way to handle those complex terms when we move from theory to what we would actually try to run on real quantum error correction systems.

Kai: The implication is that this structure isn't just a threshold effect; it has a predictable scaling behavior across different kinematic regions, which is pretty useful.

Mira: We need to be careful about those finite virtual corrections quantified by V one and V two because they introduce that quadratic logarithm in the interference, even with the better series expansion.

Lev: If we can control those coefficients, we might have a better handle on the uncertainty in our predictions for real experimental data when we're dealing with these high order calculations.

Kai: Overall, this work gives us a concrete piece of input to feed into higher order calculations at next to next to leading order at n three loop in a two h d m model.

Mira: So, the main point is that taming the analytic structure near threshold allows us to use these calculations for precision physics tests.

Lev: It’s about connecting that abstract singularity to concrete numerical scaling and error estimates for our hardware.

Kai: We're done with this one, but next up we’re looking at some work on higher order corrections in the context of Higgs production.

University of Valencia–CSIC

hep-ph

Submitted: 2026-10-13

Updated: 2026-11-09

Comments: 41 pages, 6 figures, 8 tables; threshold expansions, numerical and construction codes, and validation data included as ancillary files

Code: https://github.com/abhattacharya1992/ggAA-threshold

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

Importance score: 86/100

The gist: The two-loop amplitudes for pseudoscalar pair production in gluon fusion, gg → AA, with each pseudoscalar emitted from its own effective gluon vertex are known in closed form…

Key concepts

Threshold Expansion
This technique involves expanding the amplitude in terms of the velocity ($eta_A$) of the pseudoscalar particles near their production threshold. The expansion systematically organizes the contributions, allowing researchers to isolate specific physical singularities, such as those arising from double parton scattering (DPS), by observing which terms cancel out at different powers.
Double Parton Scattering (DPS) Singularity
This is a specific singularity in the amplitude that arises in high-energy processes like this one. The threshold expansion explicitly isolates this structure, showing that the leading logarithmic term is directly related to the DPS process. It signifies a unique kinematic feature where two partons interact simultaneously, which is crucial for understanding higher-order QCD effects.
Helicity Amplitudes
These are different components of the scattering amplitude categorized by the relative helicity (spin orientation) of the produced pseudoscalar particles. The paper shows that at two loops, these amplitudes exhibit distinct logarithmic growth patterns ($ ext{ln}3p_T$ for equal-helicity vs. $ ext{ln}2p_T$ for opposite-helicity), providing insight into the underlying dynamics of the process.
Leading Power Terms (L)
The leading power terms in the expansion are rewritten using a specific variable $L$, which is defined as a combination of logarithms. Near threshold, this variable $L$ is identified with the logarithm of the transverse momentum ($p_T$) of the pseudoscalars, up to a constant phase and power corrections. This links the theoretical expansion directly to measurable kinematic variables.

Terminology

Summary

The two-loop amplitudes for pseudoscalar pair production in gluon fusion, gg → AA, with each pseudoscalar emitted from its own effective gluon vertex are known in closed form, but large cancellations near threshold obscure their analytic structure<ref:2610. p. 0

Where the paper says it

Prepared for submission to JHEP Two-loop gg → AA amplitudes near threshold and the double parton scattering singularity Arunima Bhattacharya Theory Division, IFIC, University of Valencia–CSIC, E-46980 Paterna, Valencia, Spain E-mail: arunimab@ific. es Abstract: The two-loop amplitudes for pseudoscalar pair production in gluon fusion, gg → AA, with each pseudoscalar emitted from its own effective gluon vertex, are known in closed form, but large cancellations near threshold obscure their analytic structure.

19024#pg6>.

Threshold Expansion and Analytic Structure

The paper expands the IR-finite amplitudes in the pseudoscalar velocity βA through β6A at fixed scattering angle<ref:2610.19024#pg4>. All inverse powers of βA and the odd powers through β5A cancel, and at leading power the angular dependence enters only through ln(2pT /mA)<ref:2610. p. 0

Where the paper says it

We expand the IR-finite amplitudes in the pseudoscalar velocity βA through β 6 A at fixed scattering angle. All inverse powers of βA and the odd powers through β 5 A cancel, and at leading power the angular dependence enters only through ln(2pT /mA).

19024#pg6>. This logarithm is identified with the double parton scattering (DPS) singularity<ref:2610. p. 0
Where the paper says it

All inverse powers of βA and the odd powers through β 5 A cancel, and at leading power the angular dependence enters only through ln(2pT /mA). We identify this logarithm with the double parton scattering (DPS) singularity: with the overall constant fixed at threshold, the factorisation formula of Gaunt and Stirling reproduces the velocity dependence of its one-loop coefficient in the forward limit to 3×10−15, and a one-loop gg → HH analysis confirms the corresponding CP-even weight.

19024#pg4>. At two loops the equal-helicity amplitude grows as ln3pT, with the leading logarithmic ratio of planar N = 4 super Yang–Mills theory<ref:2610. p. 0
Where the paper says it

We identify this logarithm with the double parton scattering (DPS) singularity: with the overall constant fixed at threshold, the factorisation formula of Gaunt and Stirling reproduces the velocity dependence of its one-loop coefficient in the forward limit to 3×10−15, and a one-loop gg → HH analysis confirms the corresponding CP-even weight. At two loops the equal-helicity amplitude grows as ln3 pT , with the leading logarithmic ratio of planar N = 4 super Yang-Mills theory, while the opposite helicity amplitude, which has no one-loop DPS logarithm, grows as ln2 pT .

19024#pg6>. The opposite helicity amplitude, which has no one-loop DPS logarithm, grows as ln2pT<ref:2610. p. 0
Where the paper says it

We identify this logarithm with the double parton scattering (DPS) singularity: with the overall constant fixed at threshold, the factorisation formula of Gaunt and Stirling reproduces the velocity dependence of its one-loop coefficient in the forward limit to 3×10−15, and a one-loop gg → HH analysis confirms the corresponding CP-even weight. At two loops the equal-helicity amplitude grows as ln3 pT , with the leading logarithmic ratio of planar N = 4 super Yang-Mills theory, while the opposite helicity amplitude, which has no one-loop DPS logarithm, grows as ln2 pT .

19024#pg6>.

Leading Power and Singularity Identification

The leading power terms are rewritten in terms of L = ln βA + 1/2ln(1 − c 2) + ln 2 + iπ/2= 1/2ln(-4β 2A(1 − c 2), where the logarithm of the negative argument is taken on the branch ln(-X) = ln X + iπ for X > 0<ref:2610.19024#pg6>. Near threshold, L is therefore the logarithm of the transverse momentum of the pseudoscalars, up to a constant phase iπ/2 and power corrections<ref:2610. p. 4

Where the paper says it

After the constants that arise in the expansion are expressed in a basis of eight transcendental constants, all inverse powers of βA and the odd powers through β 5 A cancel, and every coefficient is even in cos θ, as required by Bose symmetry. At leading power, the angle enters only through the logarithm of the transverse momentum pT of the pseudoscalars.

19024#pg6>. The exact one-loop remainders already contain it, since they depend on ln(y − x) + ln(z − x) and (y − x)(z − x) = yz − 1 = (tu − m4A)/m4A by Eq. p. 3
Where the paper says it

The two-loop Type II finite remainders for gg → AA were obtained in closed form in Ref. They contain polylogarithms of weight up to four, among them Li2,2, and inverse powers of the Gram determinant tu−m4 A , which vanishes at the pair threshold.

(2.2)<ref:2610. p. 0
Where the paper says it

77 × 10−5 at the tested points; they provide this two-loop contribution to the hard function for soft-gluon resummation of the pair invariant-mass distribution near mAA = 2mA and are released with code for their numerical evaluation. Keywords: Scattering Amplitudes, Higher-Order Perturbative Calculations, Higgs Production arXiv:2610.

19024#pg6>.

Finite Virtual Corrections and Cross Section Consequences

The finite virtual corrections are quantified by dimensionless coefficients V1 and V2 relative to the tree contribution<ref:2610.19024#pg6>. The two-loop contribution to the interference with the tree amplitude gives a quadratic logarithm in its interference with the tree amplitude<ref:2610. p. 5

Where the paper says it

At leading power, the one-loop DPS logarithm does not contribute to the finite virtual interference with the tree amplitude because the two are out of phase. At the next order, the cubic logarithm of the two-loop amplitude gives a quadratic logarithm in its interference with the tree amplitude.

19024#pg6>. This leads to V2 ⊃ −7/4π 2N 2c ln2(2pT /mA)<ref:2610. p. 0
Where the paper says it

At two loops the equal-helicity amplitude grows as ln3 pT , with the leading logarithmic ratio of planar N = 4 super Yang-Mills theory, while the opposite helicity amplitude, which has no one-loop DPS logarithm, grows as ln2 pT . In the finite virtual correction, relative to the squared tree amplitude, the highest logarithm at order α 2 s is − 7 64α 2 sN2 c ln2 (2pT /mA).

19024#pg6>. The leading power expression for V1 is M(0)i βA 0 = m 2A X 3k=0 Abik L k, where L of Eq. p. 0
Where the paper says it

We expand the IR-finite amplitudes in the pseudoscalar velocity βA through β 6 A at fixed scattering angle. All inverse powers of βA and the odd powers through β 5 A cancel, and at leading power the angular dependence enters only through ln(2pT /mA).

(4.1) is the logarithm of the transverse momentum<ref:2610. p. 4
Where the paper says it

After the constants that arise in the expansion are expressed in a basis of eight transcendental constants, all inverse powers of βA and the odd powers through β 5 A cancel, and every coefficient is even in cos θ, as required by Bose symmetry. At leading power, the angle enters only through the logarithm of the transverse momentum pT of the pseudoscalars.

19024#pg6>.

Numerical Validation and Accuracy

The series through β6A reproduce the two complete remainders with maximum relative errors of 3.23 × 10−6 and 6.77 × 10−5 at βA ≤ 0. p. 0

Where the paper says it

es Abstract: The two-loop amplitudes for pseudoscalar pair production in gluon fusion, gg → AA, with each pseudoscalar emitted from its own effective gluon vertex, are known in closed form, but large cancellations near threshold obscure their analytic structure. We expand the IR-finite amplitudes in the pseudoscalar velocity βA through β 6 A at fixed scattering angle.

2<ref:2610. p. 0
Where the paper says it

77 × 10−5 at the tested points; they provide this two-loop contribution to the hard function for soft-gluon resummation of the pair invariant-mass distribution near mAA = 2mA and are released with code for their numerical evaluation. Keywords: Scattering Amplitudes, Higher-Order Perturbative Calculations, Higgs Production arXiv:2610.

19024#pg6>. The coefficients are even in cos θ, as required by Bose symmetry order by order<ref:2610. p. 4
Where the paper says it

The main results are the following: 1. After the constants that arise in the expansion are expressed in a basis of eight transcendental constants, all inverse powers of βA and the odd powers through β 5 A cancel, and every coefficient is even in cos θ, as required by Bose symmetry.

19024#pg6>. The series files are a factor of 5061 smaller than those of Ref. p. 8
Where the paper says it

The first line restores the renormalisationscale dependence absent from the released one-loop files. The second removes the overall factor 64 from the released two-loop files.

[9]<ref:2610. p. 0
Where the paper says it

77 × 10−5 at the tested points; they provide this two-loop contribution to the hard function for soft-gluon resummation of the pair invariant-mass distribution near mAA = 2mA and are released with code for their numerical evaluation. Keywords: Scattering Amplitudes, Higher-Order Perturbative Calculations, Higgs Production arXiv:2610.

19024#pg6>.

DPS Singularity and Forward Limit

The DPS singularity is probed by the threshold expansion, which isolates this structure in an explicit two-loop QCD amplitude<ref:2610.19024#pg6>. The forward limit c → 1 at fixed velocity also gives pT → 0, but the produced bosons now move along the beam<ref:2610. p. 4

Where the paper says it

At one loop its coefficient follows the factorised formula of Ref. [2] for gluon splitting and the opposite signs of the amplitudes for g +g + → A and g −g − → A: with the overall constant fixed at threshold, it reproduces the velocity dependence of the exact one-loop coefficient in the forward limit.

19024#pg6>. The one-loop DPS logarithm drops out of the tree interference because of their relative phase<ref:2610. p. 5
Where the paper says it

The origin of this double logarithm remains open. At leading power, the one-loop DPS logarithm does not contribute to the finite virtual interference with the tree amplitude because the two are out of phase.

19024#pg6>.

Two-Loop Behavior and Benchmark Comparisons

At two loops, the Jz = 0 and Jz = 2 remainders develop leading logarithms L3 and L2, respectively<ref:2610.19024#pg6>. The equal-helicity leading-logarithmic ratio agrees with the corresponding result in planar N = 4 super Yang–Mills theory<ref:2610. p. 5

Where the paper says it

At two loops the Jz = 0 amplitude, with equal gluon helicities, grows as ln3 (2pT /mA). The ratio of its leading two and one-loop logarithms agrees with the corresponding leading-logarithmic ratio in planar N = 4 super Yang–Mills theory [3].

19024#pg6>. The forward limit calculation described next indicates that the same coefficient relation persists away from threshold<ref:2610. p. 4
Where the paper says it

[2] for gluon splitting and the opposite signs of the amplitudes for g +g + → A and g −g − → A: with the overall constant fixed at threshold, it reproduces the velocity dependence of the exact one-loop coefficient in the forward limit. A one-loop calculation of gg → HH confirms the corresponding weight for a CP-even coupling.

19024#pg6>. The two-loop leading logarithms of M1 in the forward limit are therefore the one-loop DPS logarithm, with its helicity weight FA(x), multiplied by the same factor at every velocity<ref:2610. p. 0
Where the paper says it

We identify this logarithm with the double parton scattering (DPS) singularity: with the overall constant fixed at threshold, the factorisation formula of Gaunt and Stirling reproduces the velocity dependence of its one-loop coefficient in the forward limit to 3×10−15, and a one-loop gg → HH analysis confirms the corresponding CP-even weight. At two loops the equal-helicity amplitude grows as ln3 pT , with the leading logarithmic ratio of planar N = 4 super Yang-Mills theory, while the opposite helicity amplitude, which has no one-loop DPS logarithm, grows as ln2 pT .

19024#pg6>.

Application in Higher Order Calculations

The series provide the two-loop Type II input to the hard function for the pair invariant mass distribution near mAA = 2mA, at NNLO for a pseudoscalar coupled to gluons only through GG˜ and at N3LO in a 2HDM<ref:2610.19024#pg6>. The series, code for their numerical evaluation, construction code and validation records are provided with the paper [34]<ref:2610. p. 5

Where the paper says it

77×10−5 . The series and code for their numerical evaluation are provided with the paper [34].

19024#pg6>. The threshold series is recommended for 0. p. 3
Where the paper says it

The point mAA = 300 GeV corresponds to βA ≈ 0. 44; the threshold series of this paper is accurate for βA ≤ 0.

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. Enhanced calculation of two-loop Type II finite remainders using an expanded series in velocity beta A through beta 6A allows for reproduction of unexpanded amplitudes with relative errors of at most 6.77 x 10-5 for beta A less than or equal to 0.2.

  2. The improved system can explicitly identify the double parton scattering (DPS) singularity by expanding IR-finite amplitudes in velocity beta A, showing that All inverse powers of betaA and the odd powers through beta5A cancel and identifying the logarithm as the double parton scattering (DPS) singularity.

  3. The AI can perform forward limit tests on logarithmic coefficients, confirming that at leading power the angular dependence enters only through ln(2pT /mA), and verifying that at two loops, the equal-helicity amplitude grows as ln3 pT while the opposite helicity amplitude grows as ln2 pT.

  4. The system can utilize the derived coefficient relations to predict cross section behavior by evaluating D(beta) ≡ 1/piNc m2A lim c→1 d ReM(1) / d ln p2T, which tests Eq. (4.11) and shows the largest relative deviation from Eq. (4.11) was 3 x 10-15.

  5. The AI can perform flavour decomposition of the remainders to separate terms independent of nf from those proportional to nf and n2f, quantifying the cancellation by evaluating one − a0 + a1 + a2 / a0 + a1 + a2 for M1.

  6. The system can utilize the derived coefficients Abik Lk to predict the behavior of two-loop terms, showing that the cubic logarithm of the two-loop amplitude therefore gives only a quadratic logarithm in its interference with the tree amplitude.

Abstract

The two-loop amplitudes for pseudoscalar pair production in gluon fusion, gg to AA, with each pseudoscalar emitted from its own effective gluon vertex, are known in closed form, but large cancellations near threshold obscure their analytic structure. We expand the IR-finite amplitudes in the pseudoscalar velocity β A through β A 6 at fixed scattering angle. All inverse powers of β A and the odd powers through β A 5 cancel, and at leading power the angular dependence enters only through (2p T/m A). We identify this logarithm with the double parton scattering (DPS) singularity: with the overall constant fixed at threshold, the factorisation formula of Gaunt and Stirling reproduces the velocity dependence of its one-loop coefficient in the forward limit exactly, and a one-loop gg to HH analysis confirms the corresponding CP-even helicity sum. At two loops the equal-helicity amplitude grows as 3(2p T/m A), and the ratio of its leading two- and one-loop logarithms agrees with that in planar N=4 super Yang--Mills theory, while the opposite-helicity amplitude, which has no one-loop DPS logarithm, grows as 2(2p T/m A). In the finite virtual correction, relative to the squared tree amplitude, the highest logarithm at order α s squared is - 64 α s 2N c 2 2(2p T/m A). For β A at most0.2 the series reproduce the unexpanded amplitudes with relative errors of at most 6.77 times10-5 at the tested points; they provide this two-loop contribution to the hard function for soft-gluon resummation of the pair invariant-mass distribution near m AA=2m A and are released with code for their numerical evaluation.

Sources

Related papers