Daily Summary for 2026-10-20

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: It's the twentieth of October, twenty twenty-six, and this is the day's research.

Mira: 11 new papers came out today.

Kai: I'm Kai.

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

Kai: We'll take the day in one pass, then pull out the papers we're staying with.

Mira: Start us off with Toward a Physics of Deep Learning and Brains.

The day's research: Kai: Understanding topological phases helps design new materials, because it guides material design.

Mira: Researchers looked at edge states in Su-Schrieffer-Heeger chains to see their behavior.

Lev: They found that winding number changes map directly onto bulk gap closings and specific conditions on the AKMPC paper.

Kai: This establishes a fundamental link between topology and boundary behavior for classifying these systems.

Mira: In experimental physics, researchers examined low magnon damping in europium selenide films at millikelvin temperatures.

Lev: They found very low losses, with linewidth decreasing as frequency and field increase down to thirty millikelvin.

Kai: That suggests it is promising for hybrid quantum systems because of those losses.

Mira: Another area involves controlling spin using light in materials with weak intrinsic spin mixing.

Lev: They demonstrated all-optical generation of pure spin current in the altermagnet Cr two SO paper.

Kai: Infrared valley excitation combined with a THz pulse can generate nearly one hundred percent pure spin currents.

Mira: This opens a practical way to control spin using lightwave techniques in materials where intrinsic spin mixing is very low.

Kai: We explored how biological brains relate to deep neural networks in Toward a Physics of Deep Learning and Brains.

Mira: The neuronal avalanches equations apply when networks are between absorbing and active phases.

Lev: So maximal susceptibility predicts learning better than being near a true critical point?

Kai: It suggests a blueprint for engineering better network performance based on that susceptibility.

Mira: Local coherences persist unusually long in systems with quantum group symmetry, as shown in AKMPC.

Lev: AKMPC uses a Mazur bound derived from symmetry and locality to show this decay rate.

Kai: The transverse spin autocorrelation function decays at most algebraically, specifically C t 1/t squared.

Mira: In the large-anisotropy limit of the XXZ spin chain at infinite temperature, it is about sqrt 3 / 18 sqrt t.

Lev: For general chaotic models, it scales as C about t - 2/z where z is the transport exponent.

Kai: This concept of localization and phase transitions also relates to wave function behavior in disordered systems.

Mira: Studies on the Anderson transition in chiral symmetry classes show weak topology induces an intermediate quasi-localized phase.

Lev: This QL phase has a localization length that diverges only along the direction of the 1D weak topology.

Kai: The data acquisition automation proposal suggests using a time-of-flight camera and Python to measure impact crater morphology accurately.

Mira: That approach seems promising for reducing study time, but what about identifying features beyond just 3D topographic maps?

Lev: The paper on Morphological Study of Granular-Granular Impact Craters through Time-of-Flight Cameras: from Concept to Automation in Python proposes using a time-of-flight camera and a Python library to automate the study of granular impact craters.

Kai: So, it focuses on reconstruction and identification, minimizing manual measurement effort for these surfaces.

Mira: It aims for high accuracy while speeding up the analysis process significantly.

Lev: That efficiency in data handling is key when dealing with complex physical modeling requirements.

Kai: We should consider how this contrasts with methods requiring intensive manual feature extraction from raw image data.

Mira: Indeed, the automation aspect addresses that bottleneck in acquiring and processing impact crater data efficiently.

Lev: It shifts the focus from tedious measurement to algorithmic feature reconstruction using the specialized library.

Kai: That seems like a practical application of computational tools for physical modeling tasks today.

Mira: Agreed, it’s about making complex morphological studies more accessible through software implementation.

Lev: This work shows how to streamline data acquisition for granular impact craters using existing camera hardware and scripting.

Kai: So, the concrete step is leveraging Python's capabilities with a specific library for topographical mapping.

Mira: And the result is a reduction in time spent on studying these surfaces while maintaining high accuracy.

Lev: It’s an example of applying software tools directly to experimental physical data acquisition workflows.

Kai: That moves the analysis pipeline closer to real-time or rapid iterative study environments for researchers.

Mira: It makes the process scalable, which is important when dealing with a large number of impact sites.

Lev: The benefit lies in turning time-consuming image processing into an automated, reproducible procedure.

Kai: So we see a clear path toward automating complex data acquisition for physical modeling work.

Mira: Agreed, it demonstrates how software can enhance the precision and speed of morphological studies.

Lev: It provides a concrete method for acquiring and measuring morphological features of impacted surfaces reliably.

Kai: That moves the practical application forward by bridging experimental imaging with automated analysis tools.

Mira: It’s a useful demonstration of concept to automation in handling complex physical measurements effectively.

Lev: Indeed, the proposal outlines how to use Time-of-Flight cameras for this specific type of surface characterization.

Kai: So the focus is on acquiring, identifying, and measuring features by reconstructing 3D topographic maps from the camera data.

Mira: That reconstruction step is what allows for the identification of morphological features with high fidelity.

Lev: It's a systematic way to produce very accurate results for these impact crater studies using this methodology.

Kai: So, it tackles the challenge of accurately characterizing surfaces while minimizing the time required for analysis.

Mira: It’s a direct application focusing on reducing study time while ensuring high accuracy in feature measurement.

Lev: This work provides a concrete framework for automating the study of granular impact craters using this technique.

Kai: Agreed, it's about using specific tools to make complex physical characterization more routine and precise.

Mira: It shows how a concept can be translated into an automated process with measurable accuracy improvements.

Lev: The proposal sets out a clear path from the initial concept of automation to the final measurement results.

Kai: So, this work is about using specific hardware and software to achieve reliable physical measurements faster.

Mira: It’s about reducing the time spent on studying impact craters while producing very accurate results overall.

Lev: That efficiency in data handling is key when dealing with complex physical modeling requirements for these surfaces.

Kai: We should consider how this contrasts with methods requiring intensive manual feature extraction from raw image data.

Mira: Indeed, the automation aspect addresses that bottleneck in acquiring and processing impact crater data efficiently.

Lev: It shifts the focus from tedious measurement to algorithmic feature reconstruction using the specialized library.

Kai: That seems like a practical application of computational tools for physical modeling tasks today.

Mira: Agreed, it’s about making complex morphological studies more accessible through software implementation.

Lev: This work shows how to streamline data acquisition for granular impact craters using existing camera hardware and scripting.

Kai: So, the concrete step is leveraging Python's capabilities with a specific library for topographical mapping.

Mira: And the result is a reduction in time spent on studying these surfaces while maintaining high accuracy.

Lev: It’s an example of applying software tools directly to experimental physical data acquisition workflows.

Kai: That moves the analysis pipeline closer to real-time or rapid iterative study environments for researchers.

Mira: It makes the process scalable, which is important when dealing with a large number of impact sites.

Lev: The benefit lies in turning time-consuming image processing into an automated, reproducible procedure.

Kai: So we see a clear path toward automating complex data acquisition for physical modeling work.

Mira: Agreed, it demonstrates how software can enhance the precision and speed of morphological studies.

Lev: It provides a concrete method for acquiring and measuring morphological features of impacted surfaces reliably.

Kai: That moves the practical application forward by bridging experimental imaging with automated analysis tools.

Mira: It’s a useful demonstration of concept to automation in handling complex physical measurements effectively.

Lev: Indeed, the proposal outlines how to use Time-of-Flight cameras for this specific type of surface characterization.

Kai: So the focus is on acquiring, identifying, and measuring features by reconstructing 3D topographic maps from the camera data.

Mira: That reconstruction step is what allows for the identification of morphological features with high fidelity.

Lev: It's a systematic way to produce very accurate results for these impact crater studies using this methodology.

Kai: So, it tackles the challenge of accurately characterizing surfaces while minimizing the time required for analysis.

Mira: It’s a direct application focusing on reducing study time while ensuring high accuracy in feature measurement.

Lev: This work provides a concrete framework for automating the study of granular impact craters using this technique.

Kai: Agreed, it's about using specific tools to make complex physical characterization more routine and precise.

Mira: It shows how a concept can be translated into an automated process with measurable accuracy improvements.

Lev: The proposal sets out a clear path from the initial concept of automation to the final measurement results.

Kai: So, this work is about using specific hardware and software to achieve reliable physical measurements faster.

Lucky paper: 2509.22649: Kai: Okay, we're moving into Toward a Physics of Deep Learning and Brains now. This paper connects neural avalanches to deep networks, which is really interesting from a machine learning perspective because it gives us a physical language for learning dynamics.

Mira: It proposes that deep neural networks learn best when they are poised between absorbing and active phases, suggesting maximal susceptibility is a better predictor of learning than being exactly at the critical point itself.

Lev: The theoretical foundation uses non-equilibrium statistical physics to show that neuronal avalanches match the scale-free statistics seen in biological data, specifically showing power laws like P(S) ∼ S - τs and P(D) ∼ D - τd.

Kai: So they are using crackling noise theory to verify proximity to criticality, and they find Gaussian initialized networks show signatures of a genuine non-equilibrium phase transition belonging to the Barkhausen universality class.

Mira: They distinguish between two types of criticality there: the edge of chaos and the Widom-like line, which correlates with learning performance loss as sigma2b grows.

Lev: The paper shows that trainability overlaps a quasi-critical plateau, noting that control parameters giving heightened susceptibility also enable learning, and empirical results show successful training overlaps this area.

Kai: So they conclude that learning tasks require certain levels of proximity to criticality, because trainability starts to be achieved for networks initialized near the critical value sigma2w at about one.

Mira: The crackling toolkit offers operational diagnostics for training by letting pipelines track exponent-relation mismatch and distance from criticality, and standard knobs like dropout can bias training toward this quasi-critical plateau.

Lev: At the mean-field level, they show that at the critical point where sigma2w is one and sigma2b is zero the steady state signal strength q(MF)ss can take a non-trivial branch value of one/two minus one over sigma2w.

Kai: That connects the mean-field directed percolation universality class exponents, specifically beta equal to one and sigma equal to two directly to this critical behavior.

Mira: They also look at highly engineered structures like ResNets and find they exhibit signatures of criticality with exponents consistent with mean-field directed percolation.

Lev: The final conclusion is that learning is supported not by exact point criticality but by quasi-critical plateaus, meaning both brains and deep neural networks use avalanches to transmit information through processing units.

Lucky paper: 2409.03280: Kai: So we're looking at Absorbing state transitions with long-range annihilation, and the core idea is describing diffusive particles that branch and annihilate while respecting a parity constraint.

Mira: The authors introduce a continuum of unusual universality classes that interpolate between directed percolation and parity-conserving physics.

Lev: They describe this system using a mixture of deterministic and stochastic steps, including selecting a particle at random, then checking for annihilation or branching at adjacent sites with probabilities pann and pbr respectively.

Kai: That structure seems to be the mechanism driving the transition between those two main classes.

Mira: The paper finds that there is a line of universality classes with exponents that continuously interpolate between directed percolation and parity-conserving, despite having a global parity constraint.

Lev: This means we aren't just choosing between directed percolation or parity-conserving physics based on whether you have the constraint or not.

Kai: So the critical exponents themselves are changing along this line as the system parameters change.

Mira: They show that the density of particles n(t) decays as n(t) ∼ t − alpha c kappa in that critical line, where alpha c is the critical decay exponent and kappa is a parameter.

Lev: They also give us specific decay forms depending on kappa, noting that for kappa = infinity, it's n(t) about t-d/two for d=two.

Kai: What happens when we look at the short-range limit where kappa equals infinity?

Mira: In that short-range limit, particles must be neighboring to annihilate, and the decay is given by n(t) about t-one/two for d=one and kappa = zero.

Lev: But they find that for kappa < infinity, the typical interparticle distance scales on the order of r about n(t)-one/d, which changes how many annihilation events happen per unit time.

Kai: That scaling suggests a different way to think about the particle density as it evolves near criticality.

Mira: The field theory action L includes standard diffusion terms plus branching terms, but it crucially adds an annihilation term that is exponentially suppressed in particle density.

Lev: This exponential suppression comes directly from the constraint that a particle can only pair with its closest neighbor, which is why the mean-field theory relies on it.

Kai: That exponential suppression term seems like the key ingredient that allows them to match their simple scaling arguments in d two.

Mira: They also discuss how this global parity constraint exactly holds, but it loses local consequences when kappa = zero for a local subsystem.

Lev: Specifically, looking at just a local part of the system, long-range annihilation will violate the parity of particles in that subsystem even though the whole system preserves global parity.

Kai: So when you zoom in locally, the physics looks different than when you look at the whole system description.

Mira: This allows a model with a global parity constraint to be described by symmetry-free directed percolation physics rather than parity-conserving physics.

Lev: That’s a significant result because it shows how constraints can effectively change which universality class governs the local behavior.

Kai: The transition numerics show that the decay exponent alpha c is consistent with directed percolation's alpha DPc about zero point one five nine at kappa = zero.

Mira: And for kappa two it aligns with parity-conserving physics, giving a decay exponent of alpha PCc about zero point two eight five.

Lev: The exponents for zero < kappa < two seem to monotonically interpolate between those two values, which is what shows the line of critical fixed points.

Kai: That interpolation across different values of kappa gives us a very rich description of how these non-equilibrium phase transitions behave.

Mira: In the case where kappa > d, there is no longer an infrared divergence, but instead there's an ultraviolet divergence, which they regulate by the lattice spacing.

Lev: That small z-divergence in the integral is regulated by that lattice spacing, which helps keep things finite in a specific regime.

Kai: So it’s about how the spatial discretization interacts with the long-range nature of the annihilation process at high kappa.

Mira: They also predict that for every kappa two in one dimension, the decay exponent is exactly one/two except for a small discrepancy when kappa is around two point six.

Lev: This small discrepancy they mention seems to be caused by asymptotically sub-leading terms from the background long-range decay term when kappa is in that regime.

Kai: So even in this diffusion-limited regime, there's a slower transient decay at early times if you look closely at the details.

Mira: Their quantum circuit mapping uses a two-site particle number conserving block diagonal Haar matrix to create a hopping term, which allows for nonlocal annihilation processes when feedback is included.

Lev: That non-local annihilation is inspired by quantum error correction processes that often require completely non-local feedback operations when kappa equals zero.

Kai: So the classical model mimics a situation where the particle interactions are non-local in their effect on the system dynamics.

Mira: A key difference from quantum error correction is that their absorbing state dynamics don't track whether a logical error actually occurred, focusing instead on defect spawning via local branching.

Lev: The proliferation of these defects is what they suggest can induce a logical error when you try to decode something in quantum error correction.

Kai: It’s interesting how the classical model captures the idea of error induction through defect creation, rather than tracking logical information directly.

Mira: Overall, the study on Absorbing state transitions with long-range annihilation shows a continuum of critical fixed points governed by this parameter kappa.

Lucky paper: 2604.14143: Kai: Alright, we're diving into this paper titled Quantum matter is weakly entangled at low energies.

Mira: It looks like the main focus here is setting upper bounds on entanglement entropies for many-body quantum states with fixed energy expectation values.

Lev: The authors construct a relation between maximizing subsystem von Neumann entanglement entropy subject to a global energy constraint and maximizing the total thermal entropy of two independent systems at the same temperature.

Kai: So they’re linking a problem in quantum information theory directly to a classic problem in statistical mechanics involving thermal entropy.

Mira: They derive these upper bounds for both von Neumann and Rényi entanglement entropies, which they call Theorem two.

Lev: For pure states, Corollary one states that the maximum possible value of the half-system entanglement entropy is bounded by half the sum of thermal entropies.

Kai: That means they’ve found a way to use thermal entropy to constrain how much entanglement a subsystem can have.

Mira: But what about frustration-free systems, specifically those with Hamiltonians of the form H equals Pj hj?

Lev: For these frustration-free systems, ground state Schmidt ranks are upper bounded by products of the ground state degeneracies of subsystems.

Kai: If you look at the zero-temperature thermal entropies—which are logarithms of those ground state degeneracies—they must be upper bounded by subsystem surface areas for states to be area-law entangled.

Mira: That means if that scaling holds, the entanglement is confined to the boundary rather than filling the volume.

Lev: This result holds independently of any spectral gap in those frustration-free systems, which is interesting.

Kai: It suggests a structural property related to surface area dictates entanglement regardless of how gapped the system is.

Mira: What happens when you look at gapped systems at subextensive energies?

Lev: For these cases, the half-system entanglement entropy is upper bounded by two things: first, the product of the total energy in units of the gap, and second, a logarithm of system volume up to a prefactor.

Kai: So it gets more complex when you consider energy scales relative to the system's energy gap.

Mira: And for gapless systems whose heat capacities scale as powers of temperature, they establish scaling for the upper bound with both energy and system size.

Lev: In disordered systems where heat capacities scale inversely with powers of logarithms of the temperature, at subextensive energies, their upper bound scales with volume up to a multiplicative logarithmic reduction.

Kai: That logarithmic reduction suggests that disorder doesn't completely destroy the scaling behavior in these specific regimes.

Mira: Then there’s the part about nonzero energy states above the ground state where they look at systems whose heat capacities are self-averaging.

Lev: For those states at nonzero energy E minus EH zero they show that the maximum possible entanglement entropy is upper bounded by half the thermal entropy associated with that specific energy E.

Kai: So, even excited states have a bound related to their specific excitation energy level.

Mira: If the gap is gapped and E minus EH zero equals O(L alpha) where d minus one less than alpha less than d, then the entanglement entropy scales as SA of psi equals O L (d minus one ln L).

Lev: That scaling formula involves the dimension of the system, d, and it shows a logarithmic dependence on system size L.

Kai: That means for gapped systems with certain energy levels, entanglement grows logarithmically with the size of the system.

Mira: The authors also provide a lower bound construction in Appendix A to show that their upper bounds are optimal up to corrections that are subleading at large L in systems with conventional thermodynamic properties.

Lev: This simple lower bound uses a specific choice of pure state and thermal states from fictitious thermodynamics problems to establish optimality.

Kai: It’s a nice check showing that their upper bounds aren't just theoretical limits but are actually tight for many cases.

Mira: They also looked at application to free fermions in Appendix C, where they numerically show families of low-energy eigenstates whose entanglement entropies follow the scaling predicted in their upper bound fifty-three.

Lev: They found this scaling can be attributed to the presence of large numbers of long-wavelength excitations rather than states with large quantum fluctuations.

Kai: So, the physics points toward these low-energy states having a very specific excitation structure that dictates their entanglement scaling.

Lucky paper: 2509.16555: Kai: Okay, we're moving into number four of four today. We’re looking at Numerical and theoretical studies of the Anderson transition in the chiral symmetry classes with weak topology.

Mira: This paper dives deep into how 1D weak band topology affects metal-to-Anderson insulator transitions in these specific chiral symmetry classes.

Lev: The simulations confirm that this 1D weak band topology induces an intermediate quasi-localized phase between the metal and Anderson insulator phases across all three chiral symmetry classes.

Kai: So, the key finding is that this intermediate quasi-localized phase exists, and it depends on time-reversal symmetry.

Mira: The localization length in this QL phase diverges exclusively along the direction of that 1D weak topology.

Lev: They also compared universal critical exponents across these three classes to see how the nature of time-reversal symmetry influences the quantum criticality of that metal-to-QL transition.

Kai: That means different symmetries lead to different critical behaviors, which is important for material design.

Mira: They then incorporated a one-loop renormalization of the weak topological term into 2D renormalization group analysis for all three chiral classes.

Lev: This revised analysis suggests that a quasi-localized strong-coupling fixed point previously reported in the chiral unitary class turns into a conventional localized phase instead.

Kai: That’s an important shift because it means the strong coupling fixed point isn't what you initially expected to find there.

Mira: However, even with this change, in the chiral unitary and chiral symplectic classes, they still see the hallmark of the 1D weak topology through spatially anisotropic scaling.

Lev: The paper points out that this spatial anisotropy is key at the Anderson transition criticality in those two classes.

Kai: So instead of a QL strong-coupling fixed point, we are looking at anisotropic spatial scaling as the signature feature for those transitions.

Mira: They also looked at the numerical discrepancy, suggesting that the quasi-localized phase seen numerically might be an artifact of assuming spatially isotropic scaling in their finite-size scaling analysis.

Lev: The renormalization group analysis clarifies that the 1D weak topology endows the 2D Anderson transition criticality with spatially anisotropic scaling, but it doesn't support a 2D quasi-localized phase right next to the metal.

Kai: So, the model suggests that what we see numerically might be skewed by our assumption about how things scale in two dimensions.

Mira: They found that in three dee, the metal-to-QL transition has a critical exponent of nu = zero point six six plus or minus zero point zero three, which is quite different from the exponent nu = zero point eight seven eight plus or minus zero point zero zero five for the Anderson transition without the weak topology.

Lev: That difference in exponents really highlights how much time-reversal symmetry changes the nature of this critical behavior.

Kai: And they transformed these chiral NLSMs into sine-Gordon models to derive the one-loop renormalization equations for all three classes.

Mira: The metal-insulator transition is then driven by the spatial proliferation of vortex excitations of these eigenvalues in those sine-Gordon models.

Lev: The 1D weak band topology gives a complex phase factor to each vortex excitation, which interferes with the proliferation of vortex hypersurfaces that carry a finite projected volume.

Kai: That interference drives the emergence of that quasi-localized phase adjacent to the metal, which is what they are tracking in their simulations.

Mira: The RG equations for these three chiral symmetry classes show distinct fixed-point structures, and for the chiral unitary class, there's a stable weak-coupling fixed region corresponding to a 2D critical metal phase with finite conductivities.

Lev: The saddle-fixed point governing the metal-insulator transition in that same class has a conductivity ratio sigma x over sigma y approaching zero.

Kai: So, that vanishing ratio sigma x over sigma y equals zero signals that the Anderson transition criticality is characterized by this spatial anisotropy.

Mira: That implies instead of supporting a QL strong-coupling fixed point, the renormalization group points to anisotropic spatial scaling as the main signature of the 1D weak topology at that transition point.

Lev: They noted that in three dee, they observed a QL phase between conventional metal and localized phases, but in 2D with weak topology, two distinct transitions are seen.

Kai: So we have this intermediate regime being a feature that depends on the dimensionality of the system.

Mira: The paper confirms that weak 1D band topology induces an intermediate QL phase or regime between the metallic and Anderson-localized phases in chiral symmetry classes.

Lev: The inclusion of one-loop renormalization suggests that the strong-coupling fixed point is a conventional localized phase, while the saddle fixed point governing the metal-insulator transition exhibits anisotropic scaling.

Kai: This means weak topology qualitatively changes how localization happens by introducing this strong spatial anisotropy into the system dynamics.

Mira: The paper calls for further investigation of that underlying scaling structure because their current 2D finite-size scaling analysis might be affected by an isotropic scaling assumption.

Lev: They suggest that a definitive numerical confirmation of this phase requires a finite-size scaling approach that can handle generic, anisotropic spatial scaling.