Analysis of PDEs
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 14 September
Establishing global curvature estimates for $k$-convex solutions of prescribed curvature equations is a longstanding problem in fully nonlinear partial differential equations and geometric analysis. We develop a new approach that combines fractional-linear transformations with the logarithmic concavity of hyperbolic polynomials. The key step is a coercive estimate for the quadratic forms arising from the third-order terms in the maximum-principle argument. Together with a complementary concavity inequality, this estimate yields global curvature bounds for closed, strictly star-shaped $k$-convex hypersurfaces in $\mathbb{R}^{n+1}$ satisfying $\sigma_k(\kappa)=f(X,\nu)>0$, throughout the range $3\leq k<n<2k$.
Machine Learning
Task-Directed Residual AddUNet:Perfect-Reconstruction Routing for Full-Rate Representations
This paper establishes a perfect-reconstruction (PR) interpretation of AddUNet and its full-rate realization, and introduces a Residual Full-Rate PR architecture for task-directed representation learning. The survivor–skip structure of a constrained additive U-Net is shown to be exactly equivalent to a critically sampled multirate PR filter bank. The full-rate formulation removes the complementary-subband restrictions of the critically sampled system while preserving PR. A Residual Full-Rate PR architecture is then proposed to progressively route task-irrelevant, nuisance, or redundant structure away from the task-facing survivor while retaining the routed information explicitly. Exact reconstruction is guaranteed for arbitrary shape-compatible linear or nonlinear routing operators, without requiring invertibility, a matched synthesis bank, reconstruction loss, or learned decoder. The resulting architecture decouples representation design from reconstruction design: conservation is structural, while learning is devoted to task-directed routing. The same formulation identifies an identity-shortcut ResNet with its residual output retained as a full-rate PR system. Experiments verify exact single-channel routing of linearly separable factors to machine precision. On TIMIT, the proposed front-end improves test PER from $28.60\pm2.09\%$ to $25.76\pm0.41\%$ with the recognizer and training protocol held fixed, while maintaining exact reconstruction. Speaker probing further shows that structural conservation does not itself imply task-specific invariance.
Read the full abstract
Nuclear Theory
Neutron Skin Effects on Particle Emission in Heavy-Ion Collisions: A Topic Review with Astrophysical and Nuclear Structure Connections
The neutron skin, defined by the difference between neutron and proton root-mean-square radii, is a characteristic manifestation of isospin asymmetry and an important probe of the isovector nuclear interaction. This focused review examines how neutron skins influence particle emission and collective dynamics in heavy-ion collisions, from the Fermi-energy regime to ultra-relativistic energies. By modifying the initial neutron and proton density profiles, the neutron skin affects the isospin composition and geometry of the participant region, pre-equilibrium emission, particle production, fragment formation, and collective flow. We review neutron-to-proton and $\rm{t}/^3\rm{He}$ yield ratios, light clusters, pion ratios, bremsstrahlung photons, isoscaling and fragment momentum distributions, and neutron-proton differential flow and momentum observables, emphasizing their interplay with the symmetry energy and transport dynamics. At high energies, neutron skins also modify the initial geometry, eccentricities, multiplicities, and anisotropic flows in isobar and heavy-nucleus collisions. We discuss the challenge of disentangling these effects from deformation, surface diffuseness, shell structure, clustering, and model dependence. Broader connections to parity-violating electron scattering, dipole responses, coherent elastic neutrino-nucleus scattering, SRC-induced proton skins in momentum space, and neutron-star observables are also explored. Finally, we highlight opportunities from radioactive beams, improved collision experiments, microscopic many-body and transport calculations, and Bayesian inference. Combining multiple reaction systems and observables with complementary nuclear-structure and astrophysical information will be essential for quantitatively constraining neutron skins and the density dependence of the symmetry energy.
Read the full abstract
Computation and Language
K-Bench: a clinically calibrated benchmark for evaluating large language models in high-risk mental health conversations
People increasingly use large language models (LLMs) for mental health support, yet their safety in evolving, high-risk conversations remains poorly characterised. We developed K-Bench, a clinician-calibrated, protected benchmark evaluating 125 model configurations representing 33 base models from 14 providers across a fixed cohort of 200 multi-turn vignettes involving suicide, self-harm, domestic violence, substance misuse, and no-risk presentations. Synthetic patient conversations showed substantial distributional overlap with real human-AI conversations. A frozen GPT-4o judge achieved 94.2% exact agreement with clinician consensus across 6,751 eligible item comparisons from 151 clinician-rated transcripts. Leading models combined strong supportive conversation with combined-risk scores above 95, whereas risk exploration exposed substantial variation among lower-performing configurations. Therapeutic prompting produced configuration-specific gains concentrated among weaker models, while elevated reasoning produced no average improvement. K-Bench combines broader clinical coverage and configuration-scale comparison with a continuously updated public leaderboard whose operational test materials are protected from direct optimisation. The leaderboard is available at www.k-bench.ai.
Read the full abstract
Strongly Correlated Electrons
Chiral Antiferromagnetism from Momentum-Space Resonance in a 2D Semiconductor
Understanding the principles governing the emergence of chiral quantum phases is a fundamental challenge, not only for uncovering new mechanisms of quantum-state formation but also for realizing giant electronic responses and transport phenomena arising from chirality and topology. While Fermi-surface instabilities in metals can stabilize complex ordered states through multiple competing scattering channels, their microscopic origin is often obscured by the complexity of the underlying electronic structure, limiting the development of general microscopic design principles. Here, we introduce a complementary strategy based on the simplicity of semiconductor band extrema. Using the layered van der Waals semiconductor GdGaI, whose low-energy electronic structure consists of simple electron and hole valleys, we discover the spontaneous emergence of an intertwined chiral triple-$q$ antiferromagnetic state accompanied by a cooperative reconstruction of the electron-hole band edges, beyond the conventional expectation of a single-$q$ ground state. This collective reconstruction generates substantial momentum-space Berry curvature, giving rise to a pronounced spontaneous anomalous Hall effect despite the semiconducting character and negligible net magnetization. Remarkably, this chiral state is realized within an atomically well-defined ($\approx2a$), topologically nontrivial magnetic texture, showing that such collective quantum states can emerge at an exceptionally small length scale from a simple two-dimensional magnetic semiconductor. More broadly, our results introduce a remarkably simple design concept for chiral quantum matter: using simple semiconductor band extrema as building blocks for resonance-like interplay in momentum space, providing a route to Berry curvature, topological transport, and emergent quantum phases.
Read the full abstract
Analysis of PDEs
Bistable wavefronts in the Gurtin-MacCamy population model
In this paper we prove the existence of monotone traveling wave solutions for an age-structured population model with a monotone and bistable nonlinearity. We formulate the problem in terms of a discrete-time recurrence $u_{n+1} = Q[u_n]$, with operator $Q$ encoding the role of the speed into its definition. For each speed, we prove the existence of a traveling wave by the general theory of monotone evolution systems, and from appropriate estimates on the speed, we deduce the existence of a traveling wave of the original problem. Also, the sign of the speed is established under an invasion condition, and the asymptotic behavior of the profile is determined by some Tauberian results.
Read the full abstract
Combinatorics
Cauchy identities for skew Ferrers shapes via RSK and keys
Let $\mu\subseteq\lambda\subseteq(m^n)$. We characterize the image under the ordinary Robinson–Schensted–Knuth correspondence of matrices supported on the skew Ferrers diagram $\lambda/\mu$. The outer boundary determines an upper bound on the right key of the insertion tableau, while the inner boundary determines a lower bound on its left key; both bounds depend on the keys of the recording tableau. This yields tableau expansions of skew Ferrers Cauchy kernels using the standard basis polynomials of Lascoux and Schützenberger, indexed by intervals in Bruhat order. The proof first treats ordinary Ferrers diagrams. Using the supremum characterization of right keys from earlier work, we follow the $\lambda$-dependent bounds through single RSK insertions. When $\lambda$ has repeated parts, these weak column bounds need not form a semistandard tableau. Strictification determines a set $\operatorname{Comp}(\lambda)$ of admissible weak compositions and, for each $\alpha\in\operatorname{Comp}(\lambda)$, a composition $\alpha^\lambda$. Ordinary RSK then gives a weight-preserving bijective realization of the expansion \[ \prod_{(i,j)\in\lambda}\frac{1}{1-x_i y_j} = \sum_{\alpha\in\operatorname{Comp}(\lambda)} \hat K_\alpha(x)K_{\alpha^\lambda}(y), \] where $\hat K_\alpha$ and $K_\alpha$ denote Demazure atoms and key polynomials, respectively. We also give a direct admissibility criterion and a parking procedure for computing $\alpha^\lambda$. After translating conventions, these agree with the admissibility condition and half-bubble-sort construction of Feigin, Khoroshkin, and Makedonskyi. The staircase and truncated-staircase identities follow as special cases. Finally, we extend the weak-bound construction to an infinite alphabet, where strictification need not exist, and derive the infinite-variable Cauchy identity for the $m$-symmetric Schur functions.
Read the full abstract
Computation and Language
Learning to Coach for Experiential Learning
Language models can learn from experience, but raw solution trajectories are often too long and noisy to provide effective guidance. In this work, we propose Learning to Coach (L2C), a framework that trains a dedicated LLM-as-a-Coach to extract actionable experiential knowledge from an actor model's previous trajectory. The actor remains frozen, while the LLM-as-a-Coach is trained to maximize a reward given by the correctness of the actor's guided response. We study two such rewards: a same-instance reward, which improves subsequent responses on the original problem, and a cross-instance reward, which elicits knowledge that transfers to other instances. Across mathematical reasoning and interactive text-games, L2C consistently outperforms self-refinement and an untrained LLM-as-a-Coach. Running experiential learning for more iterations further improves accuracy and uses additional inference compute more effectively than enlarging the actor's decoding budget. The trained LLM-as-a-Coach also transfers to out-of-distribution tasks and adapts its guidance to the specific actor it coaches.
Read the full abstract
Combinatorics
Neither simpliciality nor mutation connectivity: conjectures of Las Vergnas and Cordovil-Las Vergnas fail
We construct a simple rank-$7$ oriented matroid on $24$ elements with no simplicial tope, disproving the Las Vergnas simplex conjecture from 1980. We then show that the mutation graph on uniform oriented matroids is disconnected for infinitely many values of rank $r$ and ground-set size $n$, disproving the Cordovil–Las Vergnas conjecture from 1988.
Read the full abstract
Computation and Language
Before You Poll with LLMs: A Deliberative Diagnostic Framework
Can LLMs reason through new information like humans, or do they merely retrieve cached opinions? This is critical for silicon sampling, where LLM personas simulate public opinion at scale. Current evaluations test only whether personas hold the right opinions – a static snapshot. But opinion research increasingly depends on dynamic fidelity: whether personas update beliefs in response to new arguments, as humans do during deliberation. No existing benchmark tests this. We introduce the Deliberative Polling Diagnostic Framework, which compares human and LLM belief shifts after identical informational interventions. Grounded in deliberative polling, it surfaces failures invisible to static evaluation: models that produce plausible partisan opinions can still misrepresent how those opinions change. Applying the framework to five frontier models using data from America in One Room (526 personas, 72 questions), we find that every model fails, each in a unique manner. GPT-5.1 exhibits reversal: its personas become more hostile toward the opposing party after balanced information, while humans become less so. This reversal is selective (80% on outgroup vs. 26% on policy questions) and symmetric across partisan identities. Gemini 2.0 Flash, Claude Sonnet 4.5, and Llama 3.3 70B exhibit overshoot, shifting correctly but at 5-7x human magnitude. DeepSeek V3 exhibits rigidity with near-zero change. Targeted ablations reveal that policy content triggers these failures and that they are identity-specific: GPT-5.1 reverses on outgroup questions but overshoots on ingroup; Gemini shows the inverse. We term this signature self-sycophancy: conformity to the model's internal stereotype of the persona rather than reasoning from the information provided. Our framework offers a concrete protocol: run the deliberative diagnostic before trusting LLM personas to mimic revised beliefs.
Read the full abstract
Data Structures and Algorithms
Scalable Triangle Counting: The Threshold Algorithm
We study one-pass triangle counting on random-order edge streams. We present a remarkably simple algorithm—read edges from the stream until $Q$ triangles are observed in the prefix, then output $Q\,(m/S)^3$ where $S$ is the stopping length—and prove that, when the maximum number of triangles incident to any edge satisfies $\eta \le T^{2/3}$, this is a $(1\pm\varepsilon)$-approximation of $T$ with probability $1-\delta$ using $O(\varepsilon^{-2}\log(1/\delta)\, m/T^{1/3})$ memory. Crucially, the algorithm does not need any a priori estimate of $T$, in sharp contrast with state-of-the-art sampling-rate based algorithms (McGregor and Vorotnikova, PODS 2020; Tsourakakis et al., KDD 2009). It also does not need a prescribed memory budget: the stopping rule self-selects the prefix length and can return an estimate before reading the entire stream. The proof rests on a Schudy–Sviridenko concentration argument for an independent-edge-sampling estimator, coupled to the without-replacement prefix produced by the algorithm. On six real temporal streams, the algorithm's stopping prefix follows the predicted cube-root scaling and achieves at most $6\%$ error at a $10\%$ prefix, without using $T$. At a fixed stored-edge budget, variance-reduced reservoir samplers are often more accurate, but only after reading the entire stream. On a separate, much larger, $1.8\times10^9$-edge graph, the threshold algorithm reads $0.46\%$ of the stream and returns $3.8\%$ error, while the strongest reservoir baselines do not finish a pass within the wall-clock cap.
Read the full abstract
Networking
Proportional-Fair Resource Allocation and Dual-Threshold Early-Exit Inference for Secure Cooperative Multi-Layer Edge Intelligence
This paper proposes FREDI (Fair Resource Allocation for Edge Dual-Threshold Inference), a secure wireless edge-intelligence framework for event-triggered inference in a cooperative user equipment (UE)–edge server (ES)–cloud system. Each UE performs early-exit convolutional neural network (CNN) screening using dual confidence thresholds, while critical events are securely offloaded to an edge server for detailed classification. We formulate a proportionally-fair utility maximization problem that jointly optimizes UE–ES association, wireless and processing resources, and confidence thresholds. FREDI decomposes the problem into proportional-fair resource allocation and dual-threshold inference optimization. We prove that the detected-critical event set is set-monotone non-increasing in both thresholds, and exploit the finite empirical confidence domain for exact threshold optimization. An empirical resource–utility response envelope yields a computable global suboptimality bound and a sufficient condition for global optimality. By pre-eliminating infeasible UE–ES pairs and exactly projecting out bandwidth and transmit-power variables, the resource-allocation subproblem is reduced to a mixed-integer exponential-cone program solvable to the certified global optimality within a prescribed gap. Numerical results with early-exit MobileNetV2 and ShuffleNetV2 demonstrate near-perfect UE fairness with aggregate utility close to a Sum-Utility benchmark, reveal security-induced resource fragmentation, and demonstrate the Stage-A scalability from 6 to 144 UEs with median solving time below 0.1~s in the tested configurations.
Read the full abstract
Accelerator Physics
Simulation study of accelerator-based muography using the GeV-scale forward muon component at SHINE
Muography exploits the penetrating power of muons to image the interior of large dense objects, but cosmic-ray sources provide only ~1 {cm}^{-2} {min}^{-1} predominantly from above, limiting imaging speed and accessible geometries. Electron-driven muon production has recently been demonstrated with laser-wakefield accelerators, but shot-to-shot fluctuations hinder systematic studies required for quantitative accelerator muography. Using Geant4 Monte Carlo simulations, we model the full experimental setup at Shaft 2 of the Shanghai High repetition rate XFEL and Extreme light facility (SHINE), from a 3 GeV, 50 pC, 50 Hz commissioning electron beam interacting with a muon target through 25 m of beamline structures and a 3 m-thick concrete isolation wall. Approximately 0.28 effective reconstructed single-muon events per bunch, with residual kinetic energies below 1.2 GeV after traversing the wall, reach the downstream muography test area (~14 s^{-1}). The wall absorbs most charged background particles below ~1 GeV, while residual neutrons can be discriminated by their characteristic detector energy deposition. Scattering-tomography simulations show that $3\times 10^{5}$ effective muon events, accumulated in approximately 6 h at the commissioning rate, yield a Structural Similarity Index above 0.9, demonstrating the feasibility of quantitative accelerator-based muography using SHINE's electron-driven GeV-scale forward muon source. At the 8 GeV/50 kHz benchmark, the projected muon intensity exceeds $5 \times 10^{4}$ {\mu}/s, corresponding conservatively to approximately one muon per bunch at the detector. This exceeds the cosmic-ray flux by orders of magnitude, while SHINE's superconducting linac provides a stable, controlled platform for developing electron-on-target muography.
Read the full abstract
Chemical Physics
Exchange-Correlation Potentials and Energies from Inverse Generalized Kohn-Sham Calculations
The Kohn-Sham (KS) formulation of density functional theory (DFT) is a map from the many-electron problem to an effective single-electron problem that is governed by a local multiplicative potential. The generalized-Kohn-Sham (GKS) formalism extends it to permit any single-electron operator—nonlocal, local non-multiplicative, local multiplicative, or any combination of them. Doing so expands the scope and ease of modeling the exchange-correlation (XC) functional in DFT, which encodes the complicated many-electron interactions into a mean-field of the electron density. However, unlike KS theory, development of XC functionals in GKS theory has been hindered by the absence of corresponding exact XC potentials and energies. We present the exact XC potentials and energies for atoms and molecules by solving the inverse GKS problem, using highly accurate correlated ab initio densities. Our approach is validated across weakly and strongly correlated systems. We further examine a common, yet untested, assumption that KS and GKS correlation potentials and energies are similar, finding instead that they differ substantially in strongly correlated systems. Overall, this work offers a powerful tool to model next-generation of XC functionals within the GKS formalism of DFT.
Read the full abstract
Atomic Physics
Absolute frequency measurement of the $^{40}$Ca$^{+}$ clock transition using a GNSS link to the SI second
We report the absolute frequency measurement of the $4s$ $ ^{2}S_{1/2}\leftrightarrow 3d$ $^{2}D_{5/2}$ $^{40}$Ca$^{+}$ clock transition with respect to the SI second. To perform this measurement, a link between our laboratory in Innsbruck and the clocks realizing the Coordinated Universal Time at the Physikalisch-Technische Bundesanstalt (PTB) in Braunschweig was installed and characterized using the Global Navigation Satellite System GNSS. The comparison between our clock and the ones at PTB was done using the Precise Point Positioning technique. After the evaluation of the systematic shifts, the measured transition frequency is 411 042 129 776 400.4$\pm$0.6 Hz with a fractional uncertainty of 1.5 $\times$ 10$^{-15}$. The stability of the clock measurements was also corroborated by comparing two different calcium ion clock experiments, which share the clock laser source at our institute. Furthermore, after careful evaluation of the trap-drive induced ac magnetic fields, we estimate ac Zeeman shifts on the $D_{5/2}$ sublevels and reevaluate the Landé g-factor of the $3d$ $^{2}D_{5/2}$ level to be g$_{5/2}= 1.200329(1)$.
Read the full abstract
Quantum Physics
Limits on the atomic description of four-wave mixing
The limits of the well-established single-atom model for describing photon-pair generation via four-wave mixing in a diamond configuration in atomic ensembles are experimentally tested. Using a cold-atom source, biphotons are generated and detected through polarization analyzers that resolve the emitted light into horizontal and vertical components in the laboratory frame. The pump lasers driving the first and second excitation transitions are horizontally and vertically polarized, respectively. To predict single-photon counts and coincidence rates in all polarization channels, the first- and second-order correlation functions are calculated directly using a comprehensive model that accounts for all Zeeman sublevels of the relevant hyperfine states. Furthermore, the population dynamics with and without the re-pump laser of the magneto-optical trap are compared, revealing substantial differences in the populations of the Zeeman sublevels. This motivates the inclusion of two additional hyperfine levels and their corresponding Zeeman sublevels in the final model. The resulting density matrix is used to calculate expectation values of the far-field electric-field operators and compare them with experimental measurements over a range of pump-laser powers. Excellent agreement with the measured photon counts is obtained over most of this range. For the coincidence measurements, good agreement is found when the polarization of the photon generated by the first (second) decay is parallel to that of the first (second) pump beam. In contrast, the model consistently underestimates the experimental coincidence rates for the opposite polarization configuration. These findings indicate that effects beyond the internal level dynamics of individual atoms, most notably collective phenomena, are required to fully account for photon coincidences generated by four-wave mixing in atomic ensembles.
Read the full abstract
Quantum Physics
Instantiating Microcrypt: Obstacles and opportunities via tailored state certification
Recent work has introduced the Hamiltonian phase state (HPS) assumptions, which postulate that Hamiltonian phase states can be used to instantiate pseudorandom and one-way state generators. Additionally, it has been conjectured that these assumptions can be true, even if one-way functions do not exist. This is exciting, because if true, then the HPS assumptions provide a route to the instantiation of Microcrypt. In this work we falsify this conjecture, by proving that if the HPS assumptions are true, then one-way functions exist. While this removes the possibility of instantiating genuine Microcrypt cryptography with Hamiltonian phase states, it shows that the HPS assumptions provide novel inherently quantum assumptions for the construction of classical cryptography. Technically we achieve this via a method for the construction of one-way puzzles from one-way state generators and tailored "measure first, ask later" state certification protocols. This generalizes prior constructions of one-way puzzles from one-way state generators via classical shadows and allows us to relate properties of the one-way puzzle to properties of the state certification protocol used in the construction. Specifically, if the state certification protocol admits efficient classical post-processing then one obtains an efficiently verifiable one-way puzzle, and if the state certification protocol can be efficiently classically simulated in a certain sense, then one obtains a classical one-way puzzle, which implies one-way functions. The latter observation allows us to prove that the HPS assumptions imply one-way functions, by exploiting properties of state certification protocols for phase states. The former observation provides a new toolbox for the construction of efficiently verifiable one-way puzzles by exploiting tailored state certification protocols for pseudorandom and one-way state generators.
Read the full abstract
Game Theory
The Existence of Maximin Share Allocations is $\Delta_2^P$-complete
The maximin share (MMS) guarantee has become one of the central fairness benchmarks for allocating indivisible items. Since Kurokawa, Procaccia, and Wang [EC'14, JACM'18] showed that exact MMS allocations need not exist, a substantial literature has developed around the existence and computation of approximate MMS allocations, including the recent work of Heidari, Kaviani, Seddighin, and Shahrezaei [SODA'26]. In contrast, a fundamental complexity question posed more than a decade ago by Bouveret and Lemaître [JAAMAS'16] has remained unresolved: how hard is it to decide whether an exact MMS allocation exists? For additive valuations, Lonc and Truszczynski [JAIR'20] established membership in the complexity class $\Delta_2^P$ (also known as $P^{NP}$), but no hardness result was known. For the more general class of 2-additive valuations, Bouveret and Lemaître established NP-hardness, leaving a substantial gap to the $\Delta_2^P$ upper bound. Moreover, the complexity of MMS existence for additive and $k$-additive valuations was posed as open questions. We resolve this longstanding question by proving that deciding the existence of an MMS allocation for additive goods is $\Delta_2^P$-complete, even for five agents. Additionally, we prove weak coNP-hardness for three agents, thereby establishing a precise dichotomy with the known existence guarantee for two agents, and strong coNP-hardness when the number of agents is unrestricted. Furthermore, our strong hardness construction yields an inverse-polynomial gap in the optimal MMS approximation ratio, ruling out an FPTAS for approximating this ratio unless P=NP. Finally, we show that all these results for goods extend to the chores setting through a polynomial-time transformation that preserves MMS existence.
Read the full abstract
Robotics
Uncertainty-Guided Sparse Refinement for Action Chunking Transformer Policies
Learning chunk-based visuomotor policies for long-horizon robot manipulation remains challenging. Recent action-chunking methods have shown promising performance by predicting temporally extended action sequences. However, their failures are often dominated by prediction errors at a small number of critical timesteps rather than uniformly poor predictions across the entire action chunk, making uniform refinement inefficient and insufficiently targeted. To address this bottleneck, we propose Uncertainty-Guided Refinement (UGR), a sparse refinement framework for chunk-based visuomotor policies. Specifically, UGR follows a coarse-to-refine design: it first predicts a full action chunk, estimates per-step temporal uncertainty from the coarse hidden states, and applies residual correction only to the most uncertain timesteps selected by a binary mask. The uncertainty branch is decoupled from the coarse action predictor, enabling clean attribution of the refinement gains to uncertainty-guided correction rather than additional predictor capacity. Extensive experiments on five dual-arm manipulation tasks from the RoboTwin benchmark show that UGR achieves the best success rate on four tasks, improves over the ACT baseline by up to 13% absolute, and outperforms both full-chunk and position-agnostic block refinement in ablation studies.
Read the full abstract
Soft Condensed Matter
Collective adaptation to time-dependent targets for active Brownian particles
We consider a minimal model of active agents which collectively learn to reach a target macroscopic state via a decentralized adaptation process. Agents are modeled as active Brownian particles which randomly reorient to a given direction through tumbling events. An average target state is encoded into a reward function evaluated by each agent. Agents can tune one of the parameters of their microscopic dynamics, called `policy' (e.g., speed or reorientation direction), to optimize their reward function thanks to information exchange with neighboring agents. We describe the evolution of the policy distribution across the agent population with a kinetic theory, under the assumption that the adaptation process is slow with respect to the physical dynamics. The adaptation process is illustrated on several learning protocols with either fixed or time-dependent target velocities. A characteristic learning time $\tau$ combining teaching and mutations emerges from the dynamics. For time-oscillating targets at angular frequency $\omega$, the collective dynamics oscillates in time, with an oscillation amplitude and phase shift determined by the product $\omega \tau$. An unlocking transition to a non-adaptable state is also observed for a rotating target velocity.
Read the full abstract
Machine Learning
Per-Matrix Optimality Is Not Enough: Three-Level Optimization for Low-Rank LLM Compression
Per-matrix singular value decomposition (SVD) truncation is Eckart-Young optimal in the whitened Frobenius norm, but errors from independently compressed matrices compound through the block's nonlinear forward pass. Inspired in part by hierarchical variational optimization in quantum many-body methods, we introduce a three-level chain that widens optimization scope from individual matrices to Transformer blocks to the full model: whitened SVD~(L1), block-level joint optimization~(L2), and end-to-end language-modeling loss refinement~(L3), all from 256 calibration sequences, with no instruction or recovery data. On LLaMA-7B at 60% compression, the chain reduces WikiText-2 perplexity from 42.1 to 19.1 to 11.4. The block-level stage acts as a regularizer: skipping it worsens Penn Treebank (PTB) perplexity by 24 points, a gap that additional end-to-end training did not close in our experiments. Perplexity gains hold across 20-80% compression, five architectures up to 13B parameters, and both in-distribution and out-of-distribution benchmarks, though the cross-architecture rows use architecture-specific configurations and the ratio sweep was not run under one common protocol. With more calibration data, skipping the block-level stage becomes competitive, revealing an offline compute–data trade-off. We therefore claim improvements only in perplexity and compression fidelity; downstream accuracy remains well below the dense model.
Read the full abstract
Data Structures and Algorithms
Consistency-Robustness Tradeoffs for Online Bipartite Allocation with Multiple Stages
We study learning-augmented online bipartite allocation with multiple stages. In the $k$-stage vertex-weighted fractional bipartite matching problem, demand vertices arrive in $k$ stages, and the algorithm receives possibly inaccurate predictions of the allocation in each stage. While tight consistency-robustness tradeoffs were known for the two-stage case, no nontrivial tradeoff was known for an arbitrary number of stages. Our main result is the first consistency-robustness tradeoff for $k$-stage vertex-weighted fractional bipartite matching with predictions, for every $k\ge2$. Let $R_k=1-(1-1/k)^k$. For every $R\in[0,R_k]$, our algorithm is $R$-robust and $C_k(R)$-consistent, where $C_k(R)=k(1-R)^{1/k}+R-(k-1)$. This simultaneously recovers the known tight two-stage tradeoff and the optimal prediction-free $k$-stage competitive guarantee $R_k = C_k(R_k)$, while strictly dominating the natural randomized coin-flip baseline between these endpoints. We also present an algorithm for the classical online setting, where demands arrive one by one and the number of demands is unknown in advance. It has a consistency ratio of at least $C_\infty(R)=1+R+\ln(1-R)$ for a given robustness $R\in[0,1-1/e]$, improving the best previously known tradeoff for this problem. Finally, we extend the framework to fractional AdWords and fractional predictions. Our algorithms are based on stage-wise convex programs with carefully calibrated vertex-dependent penalties. The penalties maintain a dynamic safety reserve for each supply vertex, balancing protection against adversarial future arrivals with the ability to exploit the predicted allocation.
Read the full abstract
Cosmology and Nongalactic Astrophysics
Cosmological Signatures of Curvature-Coupled Dark Energy
We study a curvature-coupled dark energy model that can modify cosmological evolution both before recombination and during the late-time accelerated era. The model belongs to the class of scalar-tensor theories, in which a quintessence field is non-minimally coupled to the Ricci scalar. We specify the model through a shifted quartic coupling, $f(\varphi)=\alpha(\varphi^2-\varphi_{\rm today}^2)^2$, and an inverse power-law potential, $V(\varphi)=\Lambda\varphi^{-\sigma}$, where $\varphi_{\rm today}$ is a constant fixed by requiring the effective Planck mass to recover its present-day normalization. We implement the model in a modified version of $\mathtt{hi\_class}$ and compute its background and linear cosmological predictions. This specific form of non-minimal coupling allows an effective crossing of the phantom divide at late times while naturally suppressing deviations from standard gravity today and satisfying local gravity constraints. At the same time, the scalar field can modify the expansion history before recombination, shifting the acoustic scale in the direction required to alleviate the $H_0$ tension, while remaining dynamically relevant as dark energy at low redshift. For the parameter choices studied here, we find tens-of-percent deviations from $\Lambda$CDM in the expansion history, matter clustering, and metric-potential spectra. The modified evolution of the gravitational potentials leaves characteristic signatures in relativistic observables, with weak-lensing power suppressed by $\mathcal{O}(20$-$40\%)$ at low multipoles and order-unity changes in the late Integrated Sachs-Wolfe signal. Together, these results reveal a broad, scale-dependent phenomenology that motivates both a full parameter-space analysis and an extension of the present linear treatment to a dedicated non-linear $N$-body implementation.
Read the full abstract
Computational Engineering
Block-modular grey-box identification of a short-stroke seawater-pump power take-off for wave energy: loss decomposition and multi-regime validation
Physically interpretable power take-off models support wave-energy development at every stage. For the seawater-pump class, few have been identified on a real machine and validated at the system level. This paper presents a block-modular grey-box model of a short-stroke seawater pump, identified from bench data alone. The pump decomposes into chamber, relief valve, piston seal and tip check valve, each structured by component physics, and the few parameters are identified block by block from terminal pressure, displacement, force and discharged-mass measurements, with a closed-cycle mass balance that decouples the degenerate seal and valve losses and closes to within 1% of the inflow. Over the sinusoidal campaign, the pressure-energised piston seal carries 84% of the lost mass and 97% of the lost energy, and the pump delivers about four-fifths of its input as high-pressure flow. Generalisation is demonstrated rather than assumed. With parameters fixed on ramps and sinusoids, the model reproduces 360 s sea-state records in free run to about 10% of range, captures 80% of several thousand peaks within 5 bar, transfers to a re-sprung valve on cracking strokes once its crack pressure is identified, follows thousand-cycle endurance records that drift below 0.1 bar, and reproduces the force and cycle-mean absorbed power, 0.1 to 5.7 kW, to a 4% median error. A practical-identifiability and uncertainty analysis confirms the terminal data resolve the retained parameters; thirty-seven repeated records set the experimental floor. The result is a compact, physically labelled model for design, energy assessment and model-based control of seawater-pump power take-offs.
Read the full abstract
Computer Vision
TRACE: Two-Stage Detector-Response Estimation With Angular Cosine Expansion for Ring Artifact Correction in Photon-Counting CT
Detector response nonuniformity introduces systematic projection errors and ring artifacts in photon-counting detector computed tomography (PCD-CT). In measured PCD-CT data, residual stripe amplitudes vary slowly with projection angle, which fixed-bias models cannot adequately capture. We propose TRACE, a two-stage unsupervised sinogram decomposition method for estimating and correcting these response-related errors. TRACE represents stripes as a fixed bias plus low-order discrete cosine transform (DCT) components, using a small number of coefficients to describe angular variations at each detector element. A learnable analysis–synthesis architecture represents the ideal projections, while two-stage optimization separates them from fixed and then dynamic stripes. An angular-gradient soft orthogonality constraint suppresses correlated variations within the shared DCT gradient subspace, reducing the leakage of object structures into the artifact estimate. All parameters are optimized directly on the measured sinogram without paired training data. Experiments on measured QRM mouse phantom and porcine trotter data show that TRACE suppresses ring artifacts and improves image uniformity while preserving edge sharpness, soft-tissue texture, and trabecular detail.
Read the full abstract
Computer Vision
Integrating Multi-view Multi-light Surface Reconstruction into Cultural Heritage Workflows
Cultural heritage documentation increasingly relies on image-based 3D surface reconstruction, with photogrammetry software making such workflows accessible to archaeologists, conservators, and heritage technicians. These tools have been successful for conventional multi-view acquisition, but they do not routinely exploit richer multi-view, multi-light data, despite its potential for improving fine-scale surface reconstruction. This limitation is particularly relevant in heritage contexts, where controlled-light acquisition devices such as RTI domes are already used to capture illumination-varying image sets. The challenge is therefore to connect these existing acquisition practices with recent computer vision methods in a form that can be used within operational heritage workflows. In this work, we address this need by integrating state-of-the-art components from computer vision for multi-view, multi-light surface reconstruction into Meshroom, an open-source photogrammetry framework. Rather than proposing a new reconstruction algorithm, our contribution is to assemble and expose existing advanced methods, namely a complete photometric stereo ecosystem (calibrated, self-calibrated and universal), automatic object masking, and multi-view normal-and-reflectance integration, within a usable heritage-oriented workflow. The proposed system thus provides an intermediate software layer between computer vision research code and practical cultural heritage applications, making recent techniques easier to use and evaluate.
Read the full abstract
Probability
A universality result for the critical 2d stochastic heat flow
For a large class of discrete- and continuous-time models satisfying a linear flow property, we prove convergence to the critical $2d$ stochastic heat flow under diffusive scaling of space and time. Convergence is in the sense of finite-dimensional distributions in the space of measure-valued stochastic flows. The class of discrete models we consider includes random walks in space-time random environments with finite-range jumps, and directed polymer models with finite-range spatial correlations. The class of continuum models we consider includes several distinct types of linear-multiplicative stochastic PDEs whose driving noise is Gaussian with a smooth and compactly supported covariance kernel.
Read the full abstract
Numerical Analysis
Superconvergence of Centered Finite Difference Approximations
The finite difference (FD) method is commonly used to approximate derivatives of smooth functions, with accuracy typically determined by stencil size and derivative order. However, certain centered stencils exhibit unexpectedly higher accuracy, a phenomenon known as superconvergence, which has been observed in practice but lacks rigorous explanation. We present a mathematical framework for superconvergence in centered FD approximations based on Taylor expansions of the truncation error and the resulting linear system for the FD coefficients. By analyzing symmetry properties of these coefficients and their interaction with the parity of the derivative order, we identify conditions under which higher-order error terms cancel. We show that superconvergence occurs for odd-order derivatives with even centered stencils and for even-order derivatives with odd centered stencils, while no superconvergence occurs for even derivatives with even centered stencils. Numerical experiments in MATLAB confirm the predicted convergence rates.
Read the full abstract
Computation and Language
CiteGuard-RAG: A Validation-Centered AI System for Evidence-Grounded Question Answering
Retrieval-augmented generation (RAG) can improve access to complex information; however, retrieving evidence alone does not ensure that answers are grounded, citation-valid, or appropriately refused. This paper introduces CiteGuard-RAG, a validation-centered AI system for evidence-grounded question answering. The system integrates hybrid semantic-lexical retrieval, citation-constrained generation, sentence-level grounding validation, and single-pass regeneration. Validation is used at runtime to determine whether a candidate answer should be accepted, refused, or regenerated before final delivery. CiteGuard-RAG is evaluated on 400 questions across a controlled housing-law dataset, PrivacyQA, and CUAD. In the controlled evaluation, it achieves 99.1% retrieval accuracy, 98.3% grounded-answer accuracy, and 98.3% citation validity, with no validation-detected hallucinations. Ablation results show that grounded-answer accuracy drops sharply when validation is removed, even when retrieval accuracy remains unchanged. External evaluation shows that while citation validity remains strong, evidence utilization, span alignment, and refusal calibration become harder under domain shift. These findings indicate that trustworthy RAG systems require explicit validation between retrieval and final answer delivery. CiteGuard-RAG provides a practical architecture for linking retrieval, generation, citation checking, abstention, and regeneration in high-stakes information access.
Read the full abstract
Quantum Physics
Higher-order quantum thermodynamics: equilibrium and causal structure
Quantum thermodynamics is traditionally formulated as a theory of equilibrium states and state transformations. Recent advances in higher-order quantum transformations, which describe physical scenarios beyond states and channels and provide a systematic framework for causal order, raise the question of how equilibrium should be defined and preserved in this more general setting. Starting from the Gibbs state as the unique equilibrium state, we identify equilibrium preservation as its natural higher-order extension. We show that, while this principle can in general give rise to distinct classes of transformations, all such distinctions disappear when equilibrium preservation is required completely, namely under arbitrary ancillary extensions. Remarkably, every transformation satisfying this condition is causally ordered, making causal order an emergent consequence of thermodynamic equilibrium. We establish this result for arbitrary higher-order maps and use the resulting framework to introduce free-energy-like quantities for quantum channels. Our findings reveal a fundamental connection between thermodynamic equilibrium and causal structure, providing a foundation for a fully fledged theory of higher-order quantum thermodynamics.