Finite temperature Fredholm determinants for dynamical correlation functions
summary
The gist
We study the long-time, large-distance asymptotics of finite-temperature dynamical Fredholm determinants using effective form factors This method expresses these asymptotics as a functional of an…
In short
The study uses effective form factors to find long-time, large-distance asymptotics of finite-temperature dynamical Fredholm determinants. This method determines a transition profile using parabolic cylinder functions and fixes the normalization constant by linking it to the high-temperature Gaussian beta ensemble. The leading exponential asymptotics recover generalized hydrodynamic predictions for hard rods.
Key concepts
- Effective Form Factor (EFF)
- This approach replaces original form factors with a solvable free-fermionic family. By matching its phase shift and weight to thermodynamic data, it allows the auxiliary form factor series to be summed explicitly, leaving a functional dependent only on an effective phase shift.
- Transition Profile
- This profile is determined by solving sewing conditions in a region around a quadratic saddle point. It is expressed using parabolic cylinder functions and reduces to a complex error function when the jump size is small, providing crucial information about the asymptotic behavior.
- Generalized Hydrodynamics (GHD)
- The leading exponential asymptotics of the determinant allow recovery of GHD predictions for hard rods. This involves projecting the structure factor onto modes propagating at effective velocities, weighted by occupation variance and dressed particle charge, matching numerical results.
- Parabolic Cylinder Functions
- These special functions are used to express the transition profile between the two outer branches of the determinant's asymptotic behavior. They play a key role in connecting the local profile to the high-temperature Gaussian beta ensemble's partition function.
Terminology used across episodes
This episode discusses
- Finite temperature Fredholm determinants for dynamical correlation functions · Paper Radio
- Dynamical structure factor of one-dimensional Bose gases: experimental signatures of beyond-Luttinger liquid physics
- NMR relaxation in Ising spin chains
- Evolution of Entanglement Entropy in One-Dimensional Systems
- Time-evolution methods for matrix-product states
- Emergent hydrodynamics in integrable quantum systems out of equilibrium
- Transport in Out-of-Equilibrium XXZ Chains: Exact Profiles of Charges and Currents
- Exact large-scale correlations in integrable systems out of equilibrium
- Riemann-Hilbert approach to a generalized sine kernel and applications
- Low temperature relaxational dynamics of the Ising chain in a transverse field
- Universal relaxational dynamics near two-dimensional quantum-critical points
- Low temperature correlation functions in integrable models: Derivation of the large distance and time asymptotics from the form factor expansion
- Finite temperature and quench dynamics in the Transverse Field Ising Model from form factor expansions
- Dynamical correlation functions in the Ising field theory
- Form factor approach to the asymptotic behavior of correlation functions in critical models
- Microscopic approach to a class of 1D quantum critical models
- Form factor approach to dynamical correlation functions in critical models
- One-Dimensional Quantum Liquids: Beyond the Luttinger Liquid Paradigm
- Effective free-fermionic form factors and the XY spin chain
- Large time and long distance asymptotics of the thermal correlators of the impenetrable anyonic lattice gas
- On the Dynamics of Free-Fermionic Tau-Functions at Finite Temperature
The paper
Finite temperature Fredholm determinants for dynamical correlation functions · Read on arXiv
Oleksandr Gamayun
London Institute for Mathematical Sciences · Royal Institution
We study the long-time, large-distance asymptotics of finite-temperature dynamical Fredholm determinants using effective form factors. The method expresses these asymptotics as a functional of an effective phase shift. We determine its two branches on either side of a quadratic stationary point and the transition profile joining them. The profile is expressed through parabolic-cylinder functions. Its connection to the high-temperature Gaussian β ensemble fixes the finite normalization, in agreement with an independent Riemann--Hilbert analysis. We also show that the leading exponential asymptotics suffice to recover the Euler-scale prediction of generalized hydrodynamics for the dynamical structure factor of hard rods.
Transcript
Introduction to the show: ident: Mathematical Physics. The day in mathematical physics: rigorous models of physical systems.
Kai: We're looking at the paper: "Finite temperature Fredholm determinants for dynamical correlation functions".
Mira: Here's the gist of the paper. We study the long-time, large-distance asymptotics of finite-temperature dynamical Fredholm determinants using effective form factors This method expresses these asymptotics as a functional of an effective phase shift We determine its two branches on either side of a quadratic stationary point and the transition profile joining them The profile is expressed through parabolic cylinder functions Its connection to the high-temperature Gaussian β ensemble fixes the finite normalization, in agreement with an independent Riemann–Hilbert analysis We also show that the leading exponential asymptotics suffice to recover the Euler-scale prediction of generalized hydrodynamics for the dynamical structure factor of hard rods.
Kai: The Methodology The effective form factor (EFF) approach is introduced to extend control over correlation functions in states with finite entropy Its basic idea is to replace original form factors by a solvable free-fermionic family whose phase shift and weight are chosen to match the relevant thermodynamic data.
Mira: The auxiliary form factor series can then be summed explicitly, leaving a functional of the effective phase shift.
Kai: For static determinants, this construction has a direct interpretation through an exactly solvable auxiliary Riemann–Hilbert problem In the dynamical case, matching requires joining outer solutions within a neighbourhood of width t −1/2 for a quadratic saddle.
Title and authors: Kai: So we're looking at this paper today, "Finite temperature Fredholm determinants for dynamical correlation functions". It tackles how to get these long-time, large-distance asymptotics for finite temperature dynamical Fredholm determinants using something called effective form factors.
Mira: It’s about expressing those asymptotics as a functional of an effective phase shift. That sounds like they're trying to simplify a really complex problem by boiling it down to just one thing they can calculate directly.
Lev: From an engineering standpoint, I wonder what the fidelity looks like when you try to build something that relies on matching these things up. It’s not just theory; it has to work on real hardware, right?
Kai: Exactly, Lev. The paper says they determine two branches on either side of a quadratic stationary point and then figure out the transition profile joining those two branches. They use parabolic cylinder functions to describe that profile.
Mira: Parabolic cylinder functions are interesting because they connect this problem to the high-temperature Gaussian beta ensemble. That connection is what fixes the finite normalization constant, which they agree with some independent Riemann–Hilbert analysis for static determinants >
Lev: So, if we're in the dynamical case—the real-time stuff—what does that matching part look like? How do you handle the local feature they mention?
Kai: They formulate and solve sewing conditions in a neighborhood of width t minus one over two around that quadratic saddle point. This results in a Riccati equation for the local resolvent, which is similar to what you see in beta ensembles >
Mira: That profile is expressed through parabolic cylinder functions, and it actually reduces to a complex error function when the jump size is small >
Lev: I need to know what that means for running simulations. If we use this profile, how much overhead are we talking about? Does it make the calculation tractable on a real computer?
Kai: The derivative of that profile becomes the Gauss–Wigner crossover density of the high-temperature Gaussian beta ensemble after a contour rotation >
Mira: And that connection fixes a local constant that was previously undetermined, and it gives us something called the Barnes factor, which is two pi delta over two G of one minus delta >
Lev: That sounds like they're nailing down some constants. If we can fix those things, does that mean we can actually get reliable numerical results for the dynamics?
Title and authors: Kai: The finite normalization depends on that transition profile, and it wouldn't be fixed if you chose the profile arbitrarily >
Mira: The paper says they determine this normalization constant C zero by evaluating a variational principle involving integrals needed to find C zero in equation four point one >
Lev: Evaluating integrals like that sounds computationally intensive. What is the actual form of that constant, and how does it scale? Does it blow up as we look at larger systems?
Kai: The evaluation yields C zero as delta Z R s squared p of s ds plus delta squared ZZ R squared p of s times p of s prime times the logarithm of the difference between those two points >
Mira: That looks like it’s related to the partition function of the Gaussian beta ensemble, which they define as Z N beta equals two pi N over two Y n j equal one gamma one plus j beta over two gamma one plus beta over two >
Lev: So you're connecting this to a known ensemble. If that connection holds up, it gives us confidence in the result. What about the connection to hydrodynamics they mention? Does that just mean the leading asymptotic behavior is right?
Kai: They show that only the leading exponential asymptotics are needed to recover the Euler-scale prediction of generalized hydrodynamics for the dynamical structure factor of hard rods >
Mira: That means we don't need to calculate everything perfectly if we only care about long times and big distances. The mode with rapidity k propagates at its effective velocity, which is calculated as k over the square root of K, where K is related to the density >
Lev: So for someone just looking at the resulting correlation function, what does that actually tell them? Does it give them a shortcut around brute-force calculation?
Kai: The generalized hydrodynamics projection formula reads Sc x t equals delta x times Z summed over k of rho p of k times one minus rho of k times one over one dr of k squared plus v eff of k times t > <ref:2610.18750#pg1>
Mira: That formula is what they get from the asymptotic form, and it reproduces the Euler-scale prediction for a homogeneous state >
Lev: If we use this projection formula, we don't even need to worry about that complicated finite prefactor C zero if we only care about that leading connected correlation on the Euler scale >
Title and authors: Kai: Right. The way they handle the jump-dependent contribution locally, which depends only on the discontinuity, allows them to determine that finite prefactor completely because it’s local >
Mira: It's a neat way to fix the normalization constant without needing that heavy calculation for C zero directly >
Lev: So, what about where this work stops? The paper mentions extending it to other things. What are the actual limitations they state for this specific approach?
Kai: They suggest extending it to Fredholm determinants with corrections of finite rank, which can be handled using the effective resolvent or through direct effective form factor representations >
Mira: And they also look at discontinuities already in the state, like Fermi edges at zero temperature, suggesting an analogous structure involving Tricomi confluent hypergeometric functions instead of parabolic cylinder functions >
Lev: Those extensions sound ambitious. But what’s the catch there? What are the assumptions they have to make for those other things to work?
Kai: The paper mentions that you have to establish the analytic and growth assumptions used to select the local Weber solution, and then justify the contour rotation and analytic continuation in delta >
Mira: It sounds like they're pointing out that while this method is powerful for certain states, extending it requires a solid foundation in complex analysis assumptions >
Lev: So, to wrap up this paper on "Finite temperature Fredholm determinants for dynamical correlation functions," the main point is using effective form factors to pin down the finite normalization by connecting it to the Gaussian ensemble and showing that leading asymptotics match generalized hydrodynamics >
Kai: And they do this by finding that local transition profile joining those branches, which uses parabolic cylinder functions, and it eventually allows them to get a complete picture of that finite prefactor >
Mira: It’s about using these mathematical tools to find the leading connected correlation on the Euler scale without needing the exact finite prefactor for hard rods >
Lev: For me, it means if we want to model things on real hardware, we need to be careful about those matching conditions and how complex those solutions are before we can trust the asymptotic results >
Kai: Right. We'll see if this approach helps simplify the calculations for our current qubit systems later on >
The paper's summary: Kai: So this paper is about using something called effective form factors to figure out the long time, big distance behavior of dynamical correlation functions in finite temperature states.
Mira: Right, they boil that down to an effective phase shift and then use parabolic cylinder functions to map out the transition profile between two branches.
Lev: And they fix the finite normalization constant by linking it to how things behave in the high-temperature Gaussian beta ensemble.
Kai: So what this means is that instead of just getting some rough idea, we can get a precise picture of those correlation functions using these mathematical tools.
Mira: It also shows that you only need to look at the leading exponential part of the calculation to get a good approximation for generalized hydrodynamics, which is useful for things like hard rods.
Lev: That's good because it means we don't have to calculate everything perfectly if we only care about long times and big distances. It gives us a shortcut.
Kai: The core advance is that by solving the sewing conditions right around that saddle point, they can actually pin down the finite prefactor completely using these parabolic cylinder functions.
Mira: That local profile joining the branches across their discontinuity is what lets them determine that normalization constant C zero exactly, which before was just left as an unknown parameter.
Lev: Evaluating that constant involves a variational principle and integrals like delta Z R s squared p of s ds plus delta squared ZZ R squared p of s times p of s prime times the logarithm of the difference between those two points. That sounds heavy computationally.
Kai: It is heavy, but it’s necessary because they are showing how this local jump-dependent contribution, which only depends on where the jump happens, gives us that local factor in other systems with a quadratic saddle point.
Mira: This connection to the high-temperature Gaussian beta ensemble is the key; their partition function determines that prefactor, which validates their result against an independent analysis for static determinants.
Lev: So, they've managed to nail down the finite normalization without needing that massive calculation for C zero directly by connecting it to a known ensemble. That’s a solid piece of work for testing assumptions.
Kai: And on the other side, they suggest ways to extend this method to Fredholm determinants with corrections of finite rank, which can be handled using the effective resolvent or direct effective form factor representations.
Mira: They also look at other states already having discontinuities, like Fermi edges at zero temperature, suggesting we might use Tricomi confluent hypergeometric functions there instead of parabolic cylinder functions.
Lev: Those extensions sound ambitious, but they make sense if you can establish the right analytic and growth assumptions for selecting the local Weber solution, which is a big hurdle.
Kai: Anyway, so the main point is using effective form factors to get a complete picture of that finite prefactor by solving sewing conditions locally around the saddle point.
Mira: And this leads directly to recovering Euler-scale predictions from generalized hydrodynamics without needing that exact finite prefactor for hard rods.
Lev: It confirms that those leading exponential behaviors are enough to predict things like the velocity of a mode, calculated as k over the square root of K, where K is related to density.
Kai: So for someone just listening, this means we can use this projection formula to model density correlations without needing the exact finite prefactor C zero if we only care about that leading connected part on the Euler scale.
Mira: That connection is what makes it useful; it links the complex determinant math to a physical prediction from hydrodynamics.
Lev: We need to be careful because while this works for hard rods, extending it means checking those assumptions again, especially when we look at other states or higher order corrections.
The paper's improvements: Tom: So now we're looking at what this paper suggests for making it even better, beyond just the main results, right?
Kai: They suggest extending the effective form factor construction to time dependent generalized sine kernels with a momentum dependent phase >
Mira: That means they're taking the kernel itself and making it time dependent, which is a step beyond what they did before.
Lev: And matching the original and effective kernels determines the auxiliary weight and two outer branches again, but for this new setup >
Kai: The main advance here is finding that explicit local profile joining those branches across their discontinuity in a region of width t minus one over two >
Mira: That lets them finally determine the finite prefactor of the effective determinant asymptotic completely, which was a big problem before >
Lev: Since that jump-dependent contribution is local and only depends on where the jump happens, they expect the same local factor to appear in other systems with a quadratic saddle point >
Kai: They also suggest that this method can be extended to Fredholm determinants with corrections of finite rank, which you can handle using the effective resolvent or direct effective form factor representations >
Mira: That would be useful for interacting systems, where dual field representations might provide a way to formulate that matching problem >
Lev: They also look at discontinuities already present in the state, like Fermi edges at zero temperature, suggesting an analogous structure involving Tricomi confluent hypergeometric functions instead of parabolic cylinder functions >
Kai: That sounds promising because it parallels how those parabolic cylinder functions played a role in their previous Riemann-Hilbert analysis >
Mira: But they flag that you have to establish the analytic and growth assumptions used to select the local Weber solution, and then justify the contour rotation and analytic continuation in delta >
Lev: So, while these extensions sound good for modeling real systems, they require a solid foundation in complex analysis assumptions before you can trust the results >
Kai: The paper emphasizes that the connection between those local transition profiles and the high-temperature Gaussian beta ensemble is key to fixing that jump-dependent prefactor >
Mira: That same connection determines the jump dependent prefactor, which is what makes them confident that this method will work for other systems with a nondegenerate quadratic saddle >
Lev: For me, it means if we want to model things on real hardware, we need to be careful about those matching conditions and how complex those solutions are before we can trust the asymptotic results >
Conclusion: Kai: So to wrap up this discussion on "Finite temperature Fredholm determinants for dynamical correlation functions," we've seen how effective form factors let us pin down that finite prefactor using local profile matching and connecting it to the Gaussian ensemble >
Mira: Exactly, they showed that even if we only look at leading exponential terms, we can still recover generalized hydrodynamics predictions for things like hard rods without needing the full exact prefactor >
Lev: For my perspective on running this on hardware, it means you don't need to calculate that constant C zero directly if you only care about the Euler scale correlation >
Kai: It’s a powerful way to get a result, but we have to remember those assumptions about the analytic and growth behavior of the solutions they use for the local Weber equation >
Mira: Right, and while they suggest extensions for higher rank corrections or other states like Fermi edges, establishing those complex analysis assumptions is where things get tricky >
Lev: It sounds like the method is really strong here for finding those leading behaviors, but it doesn't cover everything without those careful checks on the math >
Kai: So in short, this paper gives us a way to get precise normalization for dynamical correlation functions by solving sewing conditions locally around a saddle point and linking that to known ensembles >
Mira: It changes how we approach these finite temperature problems because it provides an exact method for finding those finite prefactors which were previously just left as unknowns >
Lev: It's a solid tool for getting the leading physics right on real hardware, provided you're prepared to handle the complex math involved in selecting those local solutions >
Kai: And we’re looking at how this framework might help us model things like interacting systems with dual field representations later on >
Mira: That's where we go next, looking at how this new formalism might apply to those more complicated, interacting scenarios.
More episodes
- 2610.33984-The Genus-One B-Model Calabi-Yau/Landau-Ginzburg Correspondence
- Daily Summary for 2026-11-19
- 2510.25312-Critical temperatures and collapsing of two-dimensional Log gases
- 2602.17549-Prefactorization algebras for the conformal Laplacian: Central charge and Hilbert Fock space
- 2604.22381-Affine Supertrusses and Superbraces
- 2605.20175-Complex deformations of the circle: Group cohomology and Virasoro uniformization
- 2504.14756-Nonlinear wave superpositions and quasi-rectifiable Lie modules
- The day in conversation for 2026-11-19
- Daily Summary for 2026-11-19
- 2508.15666-Cluster Expansion and Decay of Correlations for Multidimensional Long-Range Ising Models