Landau's Leviathans

arXiv:2606.29612 · hep-th, hep-ph · Submitted 2026-06-28 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: High Energy Physics: Theory. The day in high-energy theory: gravity, strings, fields and the questions left open.

Kai: We're looking at the paper: "Landau's Leviathans".

Mira: Here's the gist of the paper. Landau singularities are determined by tracking where the Euler characteristic of Feynman integrals drops through working over finite fields <ref:2606.29612#pg6>.

Kai: This new method extracts the entire Landau locus directly by tracking the kinematic configurations at which the Euler characteristic drops, that is, where critical points escape to infinity <ref:2606.29612#pg8>.

Mira: Working over finite fields renders the underlying elimination tractable even for many-scale, multiloop integrals.

Kai: The output, subject to verified diagnostic checks, is the strict set of singularities rather than a superset.

Paper summary: Kai: So we're looking at this paper called "Landau's Leviathans" and it’s about finding the singularities in Feynman integrals directly. What they claim is they can determine where the Euler characteristic drops by working over finite fields, which makes the math much easier for these big multi-loop integrals Kai.

Mira: That sounds like a way to track where things break down in terms of those kinematic variables s, because the abstract says they track configurations where critical points escape to infinity Mira. It’s interesting that they say this method extracts the entire Landau locus directly instead of just finding a set of potential singularities Mira.

Lev: From an error correction standpoint, if you're dealing with many-scale integrals, having a method that works over finite fields to handle the elimination process is huge because it makes the algebra tractable for what we see in real hardware Lev.

Kai: Exactly. So the core idea here is computing that Euler characteristic function, which they call χ(s), and identifying where those values drop to pinpoint these Landau singularities Kai. It’s about finding where some critical points diverge as you approach a singular point s star Kai.

Mira: And they use this tracking of critical points escaping to infinity as the definition of the Landau locus, which is pretty direct, though it relies on that projection strategy onto coordinate axes for a multivariate system Mira.

Lev: I'm curious about how they handle those complex topological issues mentioned in the appendix, like checking for higher-dimensional solutions or bulk solutions at infinity Lev. That’s where the real computational overhead usually lands when you try to simulate this on actual hardware Lev.

Kai: They have a diagnostic pipeline with four tests to make sure they only return genuine singularities and rule out things like general higher-dimensional solutions or fictitious ones at infinity Kai. It sounds like they really thought about the pitfalls of this approach Kai.

Mira: I wonder if the success of this method depends heavily on those specific checks, because if you miss one condition, you might get a superset instead of the strict set they claim to return Mira. It’s a tightrope walk between computational speed and mathematical rigor Mira.

Lev: If they can verify those conditions practically, then it suggests that for bottleneck diagrams like those two-loop six-point graphs, this approach has real promise for finding what's actually there in the physics Lev. It moves past just getting a list of possible points to getting the strict set Lev.

Kai: That's what they show. They applied it to those hard examples and reproduced all known singularities from previous studies, plus found a new set for all three families Kai. So they’re saying this isn't just theoretical; it’s working on the hardest integrals we have Kai.

Mira: The implication is that we can tackle classes of Feynman integrals that were previously out of reach with current methods, which is significant because those are the ones where you usually need these kinds of new techniques Mira. It opens up a new frontier for calculating multi-loop amplitudes Mira.

Lev: For real hardware, if you were to try running this kind of elimination on a simulator, the computational framework they describe using FiniteFlow and SPQR seems like it's designed to be computationally feasible for many-scale integrals Lev. That tractability is key for any actual physics simulation Lev.

Kai: So the paper suggests that by framing the problem over finite fields, we can make this kind of multi-parameter system solvable, which is what allows them to get past the wall they were hitting before Kai. It’s about making the algebra manageable so we can see what's happening in those complex kinematic spaces Kai.

Mira: Moving on to the conclusion of "Landau's Leviathans," it’s clear that this method provides a complete diagnostic pipeline for returning all Landau singularities, subject to those testable conditions Mira. It seems they’ve put a lot of work into ensuring the results are genuine and not just artifacts Mira.

Lev: I think the title itself, "Landau's Leviathans," reflects how difficult these kinds of structures are to pin down; it suggests they are massive problems that need a new tool to handle them Lev. It’s about tackling something huge Lev.

Kai: Yeah, and since they verified their results against existing tools like SOF A one point one, it adds a layer of confidence that this isn't just some isolated discovery Kai. It’s being checked by other methods too Kai.

Mira: What I take away is that the paper successfully demonstrated a new way to find those critical points escaping to infinity, which is fundamentally what defines these singularities in the first place Mira. It's a new way of looking at where the Euler characteristic drops Mira.

Lev: So for someone who just listens to this, it means there’s a systematic way now to look at those hard multi-loop integrals and get the actual singularity structure without missing anything obvious Lev. It gives us a better map of the bad behavior in these calculations Lev.

Conclusion: Kai: So we’ve been looking at this paper called "Landau's Leviathans." It’s about finding those tricky singularities in Feynman integrals directly, using a method that works over finite fields Kai.

Mira: Yeah, it basically takes the Euler characteristic and tracks where it drops. It sounds like they are defining the Landau locus by watching where some critical points escape to infinity Mira.

Kai: That’s right. They use this technique to map out those singularities exactly rather than just guessing where they might be in kinematic space Kai.

Lev: From my side, I’m thinking about how much work that elimination process takes. If you try to run that on real hardware, the overhead for tracking those paths could be significant Lev.

Mira: The authors mention they have four diagnostic tests to make sure they aren't just getting some random artifacts instead of actual physical singularities Mira.

Kai: They verify their results against tools like SOF A one point one, which gives it some confidence that the method is sound Kai.

Lev: It’s good that they tested it on those hard bottleneck diagrams, like those two-loop six-point graphs, to see if it actually works in practice Lev.

Mira: It suggests a systematic way to tackle these multi-loop problems where traditional methods get stuck or too complex Mira.

Kai: So, this work gives us a clearer picture of where the mathematical trouble lies in these calculations Kai.

Lev: It also shows us what kind of computational machinery would be needed if we wanted to actually implement this for real hardware simulations Lev.

Mathematical Institute, University of Oxford · Higgs Centre for Theoretical Physics, University of Edinburgh · Center for Theoretical Physics, Columbia University · Institute for Advanced Study

hep-th, hep-ph

Submitted: 2026-06-28

Updated: 2026-11-05

Comments: 9 pages, 3 figures, v2: minor fixes

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

Importance score: 91/100

The gist: Landau singularities are determined by tracking where the Euler characteristic of Feynman integrals drops through working over finite fields <ref:2606.29612#pg6>.

Key concepts

Euler Characteristic $\chi(s)$
This is a function that counts the number of solutions to a system of polynomial equations derived from an integral's representation. Landau singularities occur precisely at the points in kinematic space where this characteristic drops, signaling a change in the integral's behavior.
Working over Finite Fields
The calculation is performed using arithmetic over finite fields instead of standard real numbers. This technique makes solving the underlying algebraic elimination problems tractable, even for integrals with many scales and multiple loops.
Tracking Critical Points Escape
Singularities are located by monitoring kinematic variables ($s$) where the magnitude of a solution point $|p(s)|$ diverges as $s$ approaches a specific value $s^*$. This divergence signifies that critical points are escaping to the boundary of the parameter space.
Diagnostic Pipeline
A four-test verification process is used to ensure accuracy. These tests check for various issues like higher-dimensional solutions or fictitious singularities, guaranteeing that only true Landau singularities are returned.

Terminology

Summary

Landau singularities are determined by tracking where the Euler characteristic of Feynman integrals drops through working over finite fields <ref:2606. p. 4

Where the paper says it

The method extracts the entire Landau locus directly by tracking the kinematic configurations at which the Euler characteristic drops, that is, where critical points escape to infinity. Working over finite fields renders the underlying elimination tractable even for many-scale, multiloop integrals. p. 4

Where the paper says it

The method extracts the entire Landau locus directly by tracking the kinematic configurations at which the Euler characteristic drops, that is, where critical points escape to infinity. Working over finite fields renders the underlying elimination tractable even for many-scale, multiloop integrals.

29612#pg6>. This new method extracts the entire Landau locus directly by tracking the kinematic configurations at which the Euler characteristic drops, that is, where critical points escape to infinity <ref:2606. p. 4
Where the paper says it

DISCUSSION In this Letter we have introduced a new method for determining the Landau singularities of a Feynman integral. The method extracts the entire Landau locus directly by tracking the kinematic configurations at which the Euler characteristic drops, that is, where critical points escape to infinity.

29612#pg8>. Working over finite fields renders the underlying elimination tractable even for many-scale, multiloop integrals. The output, subject to verified diagnostic checks, is the strict set of singularities rather than a superset. p. 4
Where the paper says it

Working over finite fields renders the underlying elimination tractable even for many-scale, multiloop integrals. The output, subject to verified diagnostic checks, is the strict set of singularities rather than a superset.

Methodology for Locating Singularities

The strategy for locating the singularities of Feynman integrals revolves around computing the Euler characteristic χ(s) as a function of kinematic variables s. This involves counting the number of solutions to a system of polynomial equations derived from the integral's Lee–Pomeransky representation. The Landau singularities are precisely the locations in kinematic space where χ drops, i. p. 0

Where the paper says it

In this work we turn this idea around and propose to use χ to find the locus directly (see also [27]). In this picture, the Landau singularities are precisely the locations in kinematic space where χ drops, i.

e., where some critical points “escape” to a boundary of the parametric space. p. 0
Where the paper says it

In this picture, the Landau singularities are precisely the locations in kinematic space where χ drops, i. , where some critical points “escape” to a boundary of the parametric space.

To find these locations, one tracks values of s for which a solution point p(s) diverges as s approaches s∗. p. 2
Where the paper says it

As the kinematics change towards a singular configuration, a critical point (red) moves to infinity. This behaviour manifests itself by (at least) one projection onto a coordinate axis xi for which |p| takes finite values at s and s ′ , but diverges at s ∗ .

This is achieved by projecting the solution set onto one coordinate axis at a time using elimination ideals to reduce the multivariate problem to univariate cases. p. 2
Where the paper says it

Nevertheless, one can reduce the problem to the univariate case by projecting the solution set onto one coordinate xi at a time, before repeating the univariate analysis described above. This can be achieved by computing elimination ideals: I (i) elim ≡ I ∩ Q[xi ] .

Handling Complex Topological Issues

The method employs a diagnostic pipeline consisting of four tests to ensure that only genuine singularities are returned. These tests include checking for higher-dimensional solutions, specialized higher-dimensional solutions, bulk solutions at infinity, and fictitious singularities at infinity. p. 6

Where the paper says it

2: noDegeneracyQ returns True for every sector S, ruling out specialised higher-dimensional solutions. 1: χ = P S χS holds explicitly, ruling out bulk solutions at infinity.

For instance, Type 1. p. 4
Where the paper says it

Within each family the failure is either generic or confined to a special kinematic slice s ∗ . We label the cases accordingly, as Type 1.

1 checks if dim(JS) = 0 for every sector S to rule out general higher-dimensional solutions. p. 6
Where the paper says it

Collecting the above checks: Type 1. 1: dim(JS ) = 0 for every sector S, ruling out general higher-dimensional solutions.

Type 2.1 verifies that χ = P S χS holds explicitly, ruling out bulk solutions at infinity. p. 6
Where the paper says it

2: noDegeneracyQ returns True for every sector S, ruling out specialised higher-dimensional solutions. 1: χ = P S χS holds explicitly, ruling out bulk solutions at infinity.

Application to Bottleneck Diagrams

The method is especially suited to cutting-edge bottleneck diagrams, such as the two hardest non-planar six-point two-loop graphs and the fully massive four-scale three-loop non-planar envelope. The authors apply this technique to these examples, reproducing all known singularities from previous studies and resolving a new set for all three families. p. 3

Where the paper says it

, [9, 20, 21, 27]) either stalls or produces incomplete lists of singularities for these examples. Using the method introduced above, all known singularities of [39] were reproduced and a new set of singularities was resolved for all three families.

The resulting singularity lists are collected in the GitHub repository §. p. 1
Where the paper says it

After describing the construction in detail, we put the method to the test on two bottleneck families beyond the reach of existing tools: the two hardest non-planar sixpoint diagrams at two loops, as well as the fully massive four-scale three-loop non-planar envelope, all shown in fig. The resulting singularity lists and our proof-ofconcept Mathematica implementation are collected in the GitHub repository §.

Conclusion and Scope

The new method provides a complete diagnostic pipeline that returns all Landau singularities, barring any as-yet undiscovered failure mode. The results are verified against existing tools like SOF A (1. p. 4

Where the paper says it

We have illustrated the method on bottleneck diagrams beyond the reach of existing tools, namely the two hardest non-planar six-point two-loop graphs and the nonplanar massive envelope in fig. We provide a proofof-concept implementation together with an extensive cross-check against SOF A (1.

1). The method is designed to be computationally feasible by working over finite fields, which makes the required algebra tractable for many-scale, multiloop integrals. p. 4
Where the paper says it

The method extracts the entire Landau locus directly by tracking the kinematic configurations at which the Euler characteristic drops, that is, where critical points escape to infinity. Working over finite fields renders the underlying elimination tractable even for many-scale, multiloop integrals.

The full set of singularities can be found in the repository §. p. 3
Where the paper says it

To the best of our understanding, therefore, no genuine singularity has been missed by our analysis, and we regard these lists as complete. The full set of singularities can be found in the repository §.

The results are also provided for four-dimensional external kinematics. p. 4
Where the paper says it

– 9 – [9], which operate by default under this assumption. The singularity lists can be restricted to four-dimensional external kinematics [41], using the parametrisation of section B.

Appendix Details on Subtleties

Subtleties with the sector approach can manifest in four known ways, categorized into Type 1.1 and Type 1. p. 4

Where the paper says it

Within each family the failure is either generic or confined to a special kinematic slice s ∗ . We label the cases accordingly, as Type 1.

2 failures. Type 1.2 involves a special kinematic slice s∗ where the ideal dimension jumps, which requires testing against a recomputed ideal over the ring Q(s i) to detect the singularity. p. 5
Where the paper says it

2: Specialised Higher-Dimensional Solutions. Even when dim(JS (s)) = 0 generically, the ideal dimension can jump at a special location: dim(JS (s ∗ )) ̸= 0.

Type 2.1 checks for bulk solutions at infinity by verifying that eq. p. 5
Where the paper says it

This example is provided in more detail in the repository §. 1: Bulk Solutions at Infinity.

(12) holds before proceeding to elimination step. p. 5
Where the paper says it

One verifies that eq. (12) holds before proceeding to the elimination step.

Type 2.2 concerns fictitious singularities at infinity, which are ruled out by checking if the Euler characteristic drops at s∗. p. 2
Where the paper says it

When the ν = 0 specialisation is wellbehaved, summing the sector contributions recovers the full Euler characteristic, χ = X S∈sectors χS . (12) Finding singularities then reduces to checking whether any individual χS drops from its generic value, using the same elimination strategy applied to each JS .

Additional Context on Kinematics

The results are also provided for four-dimensional external kinematics, using a specific parametrization that imposes constraints on the Mandelstam invariants. In this restricted setting, the analysis of section II shows that the two procedures agree upon direct computation of singularities. p. 7

Where the paper says it

Although specialising kinematic parameters and computing singularities do not commute in general [19], for diagrams (b, c) of fig. 2 the two procedures agree upon direct computation: (i) restricting the d-dimensional singularity list to four-dimensional kinematics, and (ii) rerunning the singularity analysis directly on GHV and adjoining the prefactor singularities f(a) = 0, produce the same set of singularities.

The prefactor singularities arising from kinematic restrictions are also genuine Landau singularities in four dimensions. p. 7
Where the paper says it

2 the two procedures agree upon direct computation: (i) restricting the d-dimensional singularity list to four-dimensional kinematics, and (ii) rerunning the singularity analysis directly on GHV and adjoining the prefactor singularities f(a) = 0, produce the same set of singularities. Apart from these prefactor singularities, every factor arising upon restriction remains genuine in the sense of a χ-drop of eq.

The full set of singularities can be restricted to four-dimensional kinematics using the parametrisation of section B. p. 4
Where the paper says it

The singularity lists can be restricted to four-dimensional external kinematics [41], using the parametrisation of section B. We find that all restricted singularities are genuine (see the discussion around eq.

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. Automated Landau Locus Determination: The AI system can determine genuine and complete sets of singularities directly from where the Euler characteristic of the associated integral drops, which is achieved by tracking kinematic configurations at which critical points 'escape' to a boundary of the parametric space.

  2. Handling Multi-scale and Multi-loop Integrals: The system can make tractable the required algebra for multi-scale integrals at the multi-loop frontier by leveraging finite field arithmetic, allowing it to work with systems generated by many (kinematic) scale polynomial systems generated by Feynman integrals.

  3. Verification of Singularity Genuineness: The AI system implements a full diagnostic pipeline that includes four tests—Type 1.1 through Type 2.2—to verify if a candidate singularity is genuine, ensuring that the output is the strict set of genuine singularities rather than a superset.

  4. Bottleneck Diagram Analysis: The improved system can resolve and find a new set of singularities for complex bottleneck diagrams, such as (near) maximally connected graphs, which were previously beyond the reach of existing tools.

  5. Four-Dimensional Kinematics Integration: The system can restrict the singularity lists to four-dimensional external kinematics by using the parametrization of section B, and it can produce results that are genuine under this restriction.

Abstract

We present a new method together with a proof-of-concept implementation for determining the Landau singularities of Feynman integrals, read off directly from where the Euler characteristic of the associated integral drops. Working over finite fields makes the requisite elimination tractable for multi-scale integrals at the multi-loop frontier. The algorithm returns the genuine and complete set of singularities, subject to a set of conditions which are practically testable. We apply these methods to classes of Feynman integrals beyond the reach of current methods, including non-planar six-point diagrams at two loops, as well as a fully massive three-loop envelope graph. Several of the newly found singularities, both in d - and 4-dimensional external kinematics, are of unexpected complexity when compared to previously known singularities for these examples.

Sources

Related papers