Peer Revenue Publish or profit

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Friday 4 September

Neural and Evolutionary Computing

What Makes a Redundant Representation Remember? Lineage Isolation, Not Masking

Memory-based evolutionary algorithms for dynamic optimization often carry a redundant second copy of the genotype and expose only one copy to the objective, on the assumption that the shielded copy accumulates information about past optima. We show this assumption is false as usually implemented, and identify the structural property that actually determines whether the shielded copy retains information. We formalize such methods as a gated dual-copy representation with two independent design axes: a gating rule deciding which copy is evaluated, and an inheritance rule deciding whether the two copies mix across generations. A ablation shows retained information is governed almost entirely by the inheritance rule (21.4 vs. 1.3 bits) and is nearly invariant to the gating rule. Per-locus independent inheritance reshuffles cross-locus structure every generation, so shielding preserves the variance of the hidden copy while destroying the pattern that constitutes a memory. Under isolated inheritance the memory effect is real: against a single-copy baseline matched for representation budget, the method gains +0.010 AUC when optima recur periodically and loses 0.078 when they drift unidirectionally—a 0.089 separation under otherwise identical settings, which excludes explanations based on added capacity. We show the readout rate is also the corruption rate, predicting and confirming an interior optimum replicated across two implementations. We report one negative result with a mechanism: dual-copy representations lower the mutational error threshold, because gated expression is a selector rather than a joint decoder and therefore provides no coding gain. Finally, we document a benchmarking hazard: on dynamic benchmarks the choice of recombination operator alone shifted our baseline by 0.062 AUC, six times the effect size under study.
Read the full abstract

Computer Vision

Learning Spatial-Spectral Refinement and Calibrating Complementary Observations for Hyperspectral Image Super-Resolution

Hyperspectral and multispectral image fusion (HMIF) aims to reconstruct a high-resolution hyperspectral image (HR-HSI) by combining the fine spatial details of a high-resolution multispectral image (HR-MSI) with the rich spectral information of a low-resolution hyperspectral image (LR-HSI). Recent advances in implicit neural representations (INRs) have enabled flexible coordinate-based modeling for HMIF; however, existing INR-based approaches may not fully capture fine-grained spatial structures and rich spectral dependencies. Moreover, the LR-HSI and HR-MSI are primarily incorporated through degradation-consistency constraints, leaving their complementary information underexploited. To address these limitations, we propose Two-Stage Reconstruction with Implicit Tensor Neural Representation (TSR-ITNR), a unified self-supervised framework integrating representation refinement and observation-guided calibration. In Stage 1, TSR-ITNR learns an implicit Tucker representation and refines its low-rank spatial coefficient tensor and spectral basis to better capture fine spatial structures and interband correlations. A fixed pretrained denoiser further provides a deep prior for the preliminary reconstruction. In Stage 2, parameter-free calibration derives complementary and noninterfering corrections from both observations to recover information insufficiently captured in Stage 1. Theoretical analysis establishes the geometry-preserving property of spectral refinement and the orthogonal complementarity of calibration. Extensive experiments on multiple benchmark datasets demonstrate strong quantitative, visual, and spectral reconstruction performance without ground-truth HR-HSI supervision. Beyond conventional reconstruction metrics, we further assess the effectiveness of TSR-ITNR using downstream semantic segmentation accuracy.
Read the full abstract

All authors Liqian Yang, Xingchi Chen, Xinfeng Gui, Xiangyong Cao, Qianxin Yi

Number Theory

Proof of a Conjecture of Cui, Gu and Tang on 18-Colored Generalized Frobenius Partitions

Recently, the study of the number of $k$-colored generalized Frobenius partitions, denoted by $c\phi_k(n)$, has witnessed renewed interest. In this paper, we investigate congruence properties of $c\phi_{16}(n)$ and $c\phi_{18}(n)$. Our main result is a proof of the conjecture of Cui, Gu, and Tang \cite{CGT25} that, for all $n\ge0$, $c\phi_{18}(3n+2)\equiv0\pmod{2187}$. The proof uses a $(p,k)$-parametrization together with $q$-series identities and dissections. We also establish congruences for $c\phi_{16}(n)$ modulo $1024$ and $2048$, and for $c\phi_{18}(n)$ modulo $8$ and $81$.
Read the full abstract

Earth and Planetary Astrophysics

JWST Excludes Exomoons Down to 0.1 Earth Radii Around a Rocky, Temperate Exoplanet

To date, even with JWST, it has not been possible to test for exomoons as small as the Moon. Only two reported searches for exomoons around bound planets have been attempted with JWST, both of which relied on a single JWST transit. We suggest that the below expectation sensitivity to date is, in part, a product of the considerable intrinsic flexibility of the planet+moon transit model when confronted with a single event, and the presence of unanticipated systematic noise. To test this, we present a search around the rocky, temperate-zone exoplanet LP 890-9c using twelve JWST transits. We find no evidence for an exomoon but exclude $0.1R_{\oplus}$ moons to 95% confidence across the entire Hill region, representing by far the most sensitive search to date. Our limits exclude analogs to many real Solar System moons, such as Europa, Rhea and Umbriel. However, we emphasize that the close-in orbit of LP 890-9c (0.04au) would make any moons larger than ${\sim}0.1R_{\oplus}$ unlikely to survive for many Gyr due to tides. Regardless, our study firmly establishes that JWST can indeed probe down to remarkably small moons. Further, we find that even a single transit can deliver impressive limits, with the exception of one epoch that is contaminated by red noise. However, by pairing it with just one cleaner epoch, the sensitivity recovers to a level indistinguishable from any other pairing, which bodes well for a potential second JWST transit of Kepler-167e.
Read the full abstract

Robotics

Human-Human & Human-Robot Interaction Transformer (H2INT) for Robot Navigation in Dense and Uncertain Crowds

Safe robot navigation in dense crowds requires reasoning about pedestrian motion and how it may change in response to a robot. However, many learning-based approaches generate pedestrian motion independently of the robot or assume uniform reciprocity, omitting an important source of interaction uncertainty. This paper presents a Human-Human & Human-Robot Interaction Transformer (H2INT), a reinforcement learning framework that retains robot-conditioned changes in pedestrian motion during policy learning while allowing responsiveness to vary across pedestrians. Responsiveness affects the crowd dynamics when the robot is visible but is not supplied as a policy input; the policy must instead infer its consequences from robot-centered relative positions. A two-stage gated Transformer progressively encodes human-human and human-robot relations, while a recurrent policy captures their temporal evolution. A curriculum gradually reduces pedestrian responsiveness to increase interaction difficulty. Simulation experiments demonstrate improved navigation safety and robustness over representative baselines across response conditions and crowd densities, and show transfer without retraining to structurally distinct crowd-flow layouts. Ablations support the hierarchical relational encoding and gated updates. Real-robot deployment further verifies that the learned policy can operate with sparse observations in a physical environment.
Read the full abstract

All authors Ao Shen, Kaixi Chen, Shiwei Liu, Fang Deng, Chen Chen

Commutative Algebra

An Improved Upper Bound for the Perturbation Index

Let $(R,\mathfrak m)$ be a Noetherian local ring and $J$ be an arbitrary ideal of $R$. Suppose that $f_1,\ldots,f_r$ is a $J$-filter regular sequence in $R$ and $I=(f_1,\ldots,f_r)$. In this paper, we establish an improved upper bound for the perturbation index of the associated graded ring $\mathrm{gr}_J(R/I)$, refining the bound obtained by Quy and N. V. Trung.
Read the full abstract

Multiagent Systems

Online Change-point Detection for Cooperative Multi-Agent Reinforcement Learning

Cooperative multi-agent reinforcement learning (MARL) systems rely on past experience for learning coordinated behaviour, but this experience may become unreliable if the environment or task objective changes during training. In such cases, agents first need a way to recognize that the situation has changed before deciding how to adapt. This paper studies online change-point detection for cooperative MARL using reward-derived signals. We propose Patterns of Past Rewards (PPR), a lightweight algorithm-agnostic detector that smooths agents' return streams, highlights recent changes, and applies a statistical drift detector to flag significant shifts. We evaluate PPR in a custom Speaker-Listener environment based on the Multi-Agent Particle Environment under two controlled non-stationarity scenarios. Our results show a trade-off between detection speed and alarm stability. A smoothed-return baseline detects earlier but produces many repeated alarms. In contrast, applying the detector directly to raw returns often misses the shift. PPR offers a more balanced approach by limiting redundant detections while still identifying the controlled shifts. These findings highlight PPR as a lightweight, reward-based monitoring tool that enables cooperative MARL systems to reliably identify major changes during training.
Read the full abstract

Probability

The oriented Kesten–McKay law for random regular digraphs

We consider the adjacency matrix of a random directed $d$-regular graph on $N$ vertices. For fixed $d\geq 2$, we prove that the empirical eigenvalue density converges in probability to the oriented Kesten–McKay law as $N\to \infty$. The key technical input is the small-ball probability estimate for the smallest singular value. The proof combines a fixed-rank transposition argument with finite-field anticoncentration for shifted inverse compressions. We also prove a polynomial hard-edge estimate, which allows us to deduce the global law from the vanishing small-ball probability.
Read the full abstract

Computation and Language

LexFlip: A Dissociation Diagnostic for Legal Meaning Preservation Metrics

Does a simplified legal clause still say what the original said? The checks in current use cannot establish that it does: requiring an identical pair to score highest and an unrelated pair lowest moves lexical overlap and legal force together, so any monotone function of token overlap satisfies both. Our remedy is a dissociation, an item holding surface form fixed while legal force moves. We release LexFlip, 373 minimal perturbations of Quebec statutory French that reverse legal force while preserving 0.93 of the tokens, with a harness scoring metrics, regressors and prompted judges alike. The seven embedding and BERTScore metrics we test spend only 0.022 to 0.039 of their identical-to-unrelated range on such an edit, against 0.670 for bidirectional NLI, the one family the identical-pair check would disqualify. On FrJudge, against a measured human ceiling of r=0.597, a bare length feature outscores every semantic metric and has the lowest margin we measure.
Read the full abstract

Artificial Intelligence

RISE: Recursive Improvement via Self-Extrapolating Policy Distillation

On-policy distillation (OPD) provides dense, per-token supervision for language model post-training, but its effectiveness is bottlenecked by teacher quality: external teachers suffer from distribution mismatch, while self-distillation with privileged conditioning is limited by in-context learning capacity. We propose RISE (Recursive Improvement via Self-Extrapolating Policy Distillation), which constructs a synthetic teacher directly from the model's own RLVR training trajectory. By extrapolating the displacement between the current checkpoint and a trailing anchor—in parameter space or output logit space—RISE converts a sparse outcome-induced parameter update into a dense token-level target, without any external model or privileged conditioning. RISE combines RLVR and OPD in a complementary loop: outcome rewards ground the extrapolation toward correct reasoning, while the extrapolated teacher refines token-level decisions. Moreover, since the teacher is refreshed every iteration as the student improves, distillation becomes a recursive improvement mechanism rather than a one-shot compression step. Experiments spanning mathematical reasoning, multi-domain STEM, code generation, and multi-turn agentic tasks show that RISE outperforms RLVR-only training and on-policy self-distillation across all settings.
Read the full abstract

Machine Learning

Learning from VAE Errors to support ECG-based Differential Diagnosis of Myocardial Scar

Late Gadolinium Enhancement (LGE) on cardiac magnetic resonance is a key marker of myocardial scar, but its limited accessibility motivates routine ECG-based screening. We evaluated whether $\beta$-variational autoencoder (VAE)-derived ECG representations can discriminate LGE+ from LGE- cardiomyopathic patients in a local cohort of 300 subjects. We compared 32-dimensional features from the foundation ECGx.AI model with those from a shallower $\beta$-VAE trained on normal PTB-XL ECGs, evaluating downstream classification and Dynamic Time Warping (DTW)-based reconstruction errors. ECGx.AI reached an area under ROC of 0.686 with Random Forest, while the proposed $\beta$-VAE reached 0.577 with sensitivity of 0.775 with Gradient Boosting. Notably, DTW-reconstruction errors significantly differed between classes in 10 out of 12 leads according to Mann-Whitney U test and help in classification, leading to an area under ROC of 0.643 with Logistic Regression, supporting their potential as markers of scar-related ECG alterations.
Read the full abstract

All authors Shayan Sharifi, Riccardo Treu, Ilaria Gandin, Federico Garoia, Marco Merlo, Giulia Cisotto

Analysis of PDEs

Solutions to fourth order degenerate parabolic equations obtained via weighted energy dissipation

We apply the variational method of Weighted Energy Dissipation (WED) to obtain a global-in-time approximation of solutions to fourth order degenerate parabolic equations of Cahn-Hilliard type. Differently from the standard approach to the existence theory, WED induces an elliptic regularization in time, not in space. The confinement on the phase field is guaranteed already for the approximation, a priori estimates follow without modification of Lyapunov functionals, and the approximations are weakly differentiable in time — even twice in the case of linear mobility. While WED has been widely used for the construction of gradient flows in Hilbert spaces, this appears to be the first application of WED to a metric gradient flow beyond second order PDEs of Wasserstein type.
Read the full abstract

Combinatorics

On the lengths of MDS codes with a two-transitive permutation automorphism group

Let $C$ be an $[n,k]_q$ maximum distance separable (MDS) code with $4\le k\le q-3$, and suppose that it has a $2$-transitive permutation automorphism group. In this paper we show that $n\le q+1$, so the MDS conjecture holds for this class of codes.

High Energy Physics (Phenomenology)

Inelastic from the Other Side: Xenon Excitation Signals in Light of the LZ High-Recoil Event

The LUX-ZEPLIN (LZ) experiment recently reported a single event consistent with a nuclear recoil of $E_{\rm NR}\sim248$~keV in an extended-energy search. We study the accompanying inelastic-xenon channel, $\chi+Xe\to\chi+Xe^{*}$, as a complementary test of dark matter (DM) interpretations of such high-recoil events. Using the non-relativistic effective-field-theory (NREFT) framework, we compute the inelastic-xenon rates for representative scenarios including accelerated DM populations and inelastic dark matter. Their signal patterns in the S1–S2 plane are reconstructed. The nuclear de-excitation shifts these signals toward the electronic-recoil band, where they may overlap with clustered isotope backgrounds. We then propose a schematic, physics-motivated background model to estimate their detectability. For many benchmarks, the inelastic-xenon rates are $\mathcal{O}(0.1)$ of the elastic rates or below, and the projected sensitivity remains weak even at $200~{\rm tonne\,yr}$ for the scenarios investigated. Our results provide additional inelastic-xenon signal predictions for various DM scenarios and motivate improved high-energy background modeling and detector calibration for future searches.
Read the full abstract

All authors Guanhua Gu, Lingfeng Li, Shao-Song Tang, Yongheng Xu

Probability

Universal Exponent-Two Degree Laws in Range-Renewal Networks

Let an infinite sequence of independent and identically distributed random variables over a countable alphabet generate a graph by joining consecutive symbols and suppressing repeated edges. We determine the exact tail and local asymptotics of the limiting degree distributions of this range-renewal graph. If the ordered sampling probabilities satisfy \(\pi_k\in\mathrm{RV}_{-1/\gamma}\) with \(0<\gamma<1\), then the directed and undirected degree tails are asymptotic to \(\pi_k^\gamma\) and \(2^\gamma\pi_k^\gamma\), respectively, while the corresponding local masses are asymptotic to \(\pi_k^\gamma/k\) and \(2^\gamma\pi_k^\gamma/k\). Consequently, both limiting laws are regularly varying with index \(-2\), independent of \(\gamma\); forgetting edge orientation affects only the leading amplitude. The proof combines infinite-occupancy estimates for discovery times and residual unseen mass with a conditional geometric representation of inter-discovery gaps. Uniform integrability yields the tail asymptotics, whereas a geometric-smoothing argument obtains the local masses without differentiating a regularly varying tail. We also prove that deleting self-loops leaves the limiting laws unchanged. Finite-sample simulations for normalized Zipf frequencies illustrate the asymptotic result.
Read the full abstract

Artificial Intelligence

Beyond Aggregate Scores: Behavioral Correctness Assumptions for Assessing Reference-Based Automatic Evaluation Methods

Automated reference-based evaluation methods play a critical role in assessing natural language generation systems. Existing meta-evaluation primarily measures agreement with human judgments or benchmark labels, providing limited insight into evaluator behavior under controlled conditions. We introduce behavioral correctness assumptions, a complementary framework for evaluating reference-based automatic evaluation methods. We define a taxonomy of correctness-preserving and correctness-altering assumptions and operationalize them through controlled response transformations that specify expected scoring behaviors. We evaluate diverse lexical, character-level, semantic, LLM-based, and hybrid evaluators and analyze their assumption-level behavior, stability, sensitivity, repeat-run variability, configuration sensitivity, and reproducibility. Our experiments reveal distinct behavioral trade-offs across evaluation paradigms: no evaluator satisfies all proposed correctness assumptions, and evaluators with similar aggregate performance can exhibit substantially different behavioral profiles. These findings demonstrate that behavioral correctness assumptions provide diagnostic information obscured by conventional aggregate meta-evaluation.
Read the full abstract

All authors Maria Mahbub, Ashley Rice, Michael R. Munroe, Amidu Kamara, Amir Sadovnik

Programming Languages

Augur: Predicting View Serializability Violations in Relational Data Store Applications

Data stores are widely used because they provide persistence, scalability, and fault tolerance with a simple interface. However, most data store applications configure the data store to use weak isolation to achieve scalable performance, resulting in sporadic unserializable executions that are incorrect or fail. Prior work uses dynamic predictive analysis to infer violations from execution traces, but it cannot handle relational (i.e., SQL) queries with complex predicates, and it predicts executions that do not violate View Serializability. This paper introduces Augur, the first dynamic predictive program analysis that (1) supports data store applications with complex relational queries and (2) reports only executions that violate View Serializability. The evaluation demonstrates that Augur finds feasible, unserializable executions in the widely used OLTP-Bench programs and in the widely used e-commerce application Spree.
Read the full abstract

All authors Chujun Geng, Noah Charlton, Spyros Blanas, Michael D. Bond, Yang Wang

Quantum Gases

Fluctuations of a Photon Bose-Einstein Condensate Coupled to a Reservoir: Describing Coherence Properties in a Free-Energy Model

Photons are mutually nearly noninteracting particles, so thermalized photon ensembles are commonly obtained not from direct particle-particle-interactions but rather from contact with matter, which can constitute a reservoir for the photon gas. We develop a theory model for photons in a material-filled (e.g. liquid dye) optical microcavity, with the aim to study the fluctuation properties using a free-energy description for noninteracting photons coupled to a reservoir of material electronic excitations. To begin with, we use a single mode description for the condensate. For a small relative size of the material reservoir, corresponding to the canonical regime, condensate number fluctuations are small, and the derived free energy landscape takes the usual Mexican-hat shaped form such that spontaneous symmetry breaking occurs. In contrast, for a large relative size of the reservoir, corresponding to the grand canonical regime, fluctuations become as large as the average particle number. We show that the resulting free energy landscape acquires a bowl-shaped form, with a single minimum at the origin. Thus, a macroscopic occupation of the ground state (i.e., Bose-Einstein condensation) in the absence of spontaneous symmetry breaking is expected. We also provide a model for the treatment of a photon gas trapped in a box-shaped potential with spatially distributed coupling to a reservoir. The model predicts, for example, a statistically fluctuating pattern of islands with long-range coherence, resembling transient microcondensates.
Read the full abstract

All authors Martin Weitz, Andreas Redmann, Riccardo Panico, Leon Kleeblank, Kevin J.H. Peters, Frank Vewinger, Julian Schmitt

Statistics Theory

Statistical reduction before the target is known: two boundary results

Suppose that the eventual use of data is not known when the data are reduced or collected. This note considers two simple boundary cases. In a finite statistical experiment, a statistic preserves the Bayes risk for every finite later decision problem if and only if it is sufficient. Hence, when the minimal sufficient statistic is one-to-one, exact preservation of all later decision problems permits no nontrivial reduction. We then consider adaptive sampling from $m$ independent Gaussian streams when an external query specifies the coordinate to be classified only after sampling stops. Under coordinatewise error control, the optimal symmetric average sample size is exactly $m$ times the one-coordinate optimum. A change-of-measure argument gives the corresponding pointwise lower bound in terms of binary relative entropy.
Read the full abstract

Astrophysics of Galaxies

Rapid quenching and early gas depletion in the core of a galaxy protocluster at z=2.2

Understanding how galaxy evolution is affected in forming high-z overdensities is a major open question in modern astrophysics. Here, we present the average properties of the quiescent galaxy population recently identified in the core of the Spiderweb protocluster at z = 2.16. By stacking observations from the Hubble Space Telescope (HST; F160W imaging and G141 grism spectroscopy), the James Webb Space Telescope (JWST; PaBeta narrowband imaging), and the Atacama Large Millimeter/submillimeter Array (ALMA; 1.2 mm dust continuum imaging), we constrained their ensemble star formation history (SFH), residual dust-obscured star formation, and cold gas mass, respectively. The inferred average SFH indicates steady stellar mass growth, with progenitor SFRs consistent with those of massive Halpha emitters (HAEs) in the same structure. This is followed by rapid quenching, on average 500 Myr prior to observation. No significant age differences are found between active galactic nucleus (AGN) hosts and non-AGN systems. We find weak PaBeta emission for 80% of the sample, yet with dust-obscured star formation rates consistent with quiescence in all but one source (ID 443) excluded from the sample. Our galaxies remain individually and globally undetected in the dust continuum, yielding an average gas fraction fgas= Mmol/Mstar<7%, consistent with previously reported gas depletion trends with stellar mass observed in neighboring HAEs. We thus find the Spiderweb protocluster to host a sizable population of gas-depleted, massive quenched galaxies in its core, likely descending from main-sequence galaxy progenitors and building the first elements of the red sequence in this high-z forming cluster.
Read the full abstract

All authors C. D'Eugenio, H. Dannerbauer, A. Naufal, J. M. Pérez-Martínez, Y. Koyama, P. G. Pérez-González, Y. H. Zhang, T. Kodama, K. Daikuhara

Artificial Intelligence

GUT: Quantifying and Optimizing the Reasoning Uncertainty of LLMs via Graph Complexity

Recent years have witnessed great advances in the reasoning ability of Large Language Models (LLMs). However, the reasoning processes of LLMs often exhibit uncertainty, where LLMs often produce a proliferation of divergent branches at each reasoning step even when fed the same prompting inputs, and certain branches exhibit evidently incredible, even nonsensical, reasoning chains and results. In this paper, we propose the Graph-complexity-based UncerTainty (GUT) method for investigating the reasoning uncertainty of LLMs. The key idea of GUT is to characterize the potential branches of each reasoning chain with a directed acyclic graph, thereby ensuring that all potential branches are comprehensively covered within the graph space. Building upon this recognition, we further build two modules of GUT, that is, a Quantification (GUT-Q) module and an Optimization (GUT-O) module, for quantifying and reducing the reasoning uncertainty of LLMs, respectively. GUT-Q measures LLM reasoning uncertainty by approximating the reasoning space complexity with graph complexity. GUT-O implements uncertainty optimization by treating negative uncertainty as the reward function in reinforcement learning. Experimental results conducted on four LLMs and five datasets validate the effectiveness of GUT.
Read the full abstract

All authors Shuang Liang, Xin-Yu Hu, Xiang-Jun Ou, Shao-Qun Zhang

Analysis of PDEs

Asymptotic expansion of the voltage potential under small Robin perturbations of its boundary conditions

Inspired by questions related to inverse problems and shape optimization, we derive an asymptotic expansion of the voltage potential, solution to a model elliptic second-order partial differential equation, under small perturbations of its boundary conditions. More precisely, the homogeneous Dirichlet or homogeneous Neumann boundary condition in a fixed, reference configuration of the problem is replaced by a Robin boundary condition with admittance $k_\varepsilon > 0$ on a “small” subset $\omega_\varepsilon$ of the boundary of the ambient domain, vanishing at the limit $\varepsilon \to 0$. In each of these two situations, a general asymptotic formula is established for the voltage potential, which rests on minimal assumptions about the shape of the vanishing subset $\omega_\varepsilon$ and the parameter $k_\varepsilon$. The scalings of these expansions, capturing the intensity of the perturbation, are measured by new quantities called “Robin-Dirichlet” capacity or a “Robin-Neumann” capacity, which depend on the geometry of $\omega_\varepsilon$ and on the value of the parameter $k_\varepsilon$. We analyze how these quantities compare to more classical measures of the “smallness” of $\omega_\varepsilon$, such as the capacity or the Neumann capacity of $\omega_\varepsilon$, according to the behavior of $k_\varepsilon$ as $\varepsilon \to 0$.
Read the full abstract

Robotics

Temporal Tactile Encoding and Compliance for Intent-Aware Robot-to-Human Bimanual Handover

Reliable robot-to-human handover requires the robot to infer when the person is ready to receive the object and release it safely, comfortably, and at the right time. This is challenging because visual observations alone may not disambiguate clear taking intent from accidental contact, weak grasping, wrong-direction forces, or transient interactions. In this work we treat human-robot handover as an intrinsically multimodal problem. Our approach couples a VLA model with a compliance controller that reduces interaction forces during object transfer. We finetune the VLA model with human demonstrations using RGB observation, temporally encoded tactile feedback and proprioception. We evaluate the complete system in a human-subject study against two baselines: one without tactile feedback and one using tactile feedback without compliance control. We hypothesize that combining compliance and temporal tactile encoding yields the most reliable and comfortable handovers, as compliance facilitates physical interaction while tactile history captures sustained taking intent. Performance is measured through objective metrics and an ad-hoc questionnaire. The results show that the two components provide complementary benefits and substantially outperform the baselines. Project page: https://hsp-iit.github.io/temporal-tactile-bimanual-handover/.
Read the full abstract

All authors Pasquale Marra, Stefano Berti, Gabriele Mario Caddeo, Lorenzo Natale

Sound

KanAdapter: A Kolmogorov-Arnold Network-based Plug-and-Play Module for Efficient Fine-tuning of Foundation Speech Models

Fully fine-tuning self-supervised learning (SSL) speech models for downstream tasks is computationally prohibitive, and existing parameter-efficient fine-tuning approaches predominantly rely on MLP-based adapters whose fixed activation functions limit their representational expressiveness under tight parameter budgets. We propose KanAdapter, a lightweight adapter framework that replaces conventional MLP bottlenecks with Group-Rational Kolmogorov-Arnold Network (GR-KAN) modules for more expressive and parameter-efficient adaptation. Following a parallel bottleneck design, KanAdapter inserts trainable GR-KAN branches alongside frozen Transformer encoder blocks and leverages weight transfer from pre-trained MLP layers for stable initialization. Across speaker verification, speech emotion recognition, and deepfake detection, KanAdapter achieves up to 97.5% reduction in trainable parameters relative to full fine-tuning while remaining highly competitive, and consistently outperforms AdaptFormer under comparable parameter budgets. In continual learning, it yields up to 83.6% error reduction over full fine-tuning and MLP-based adapters, which we attribute to the localized nature of GR-KAN's rational activations that mitigates catastrophic forgetting. To our knowledge, this is the first work to explore KAN-based modules for parameter-efficient fine-tuning of speech foundation models.
Read the full abstract

Differential Geometry

Sophus Lie's problem on two-dimensional metrics with projective symmetries: completing the local classification

We complete the local classification, up to isometry, of $2$-dimensional pseudo-Riemannian metrics (i.e. both Riemannian and Lorentzian), admitting a projective symmetry algebra of dimension at least two. The new contribution is the treatment of the non-regular case: we obtain a complete list of mutually non-isometric normal forms in neighbourhoods of points where the action of the projective Lie symmetry algebra fails to be regular, including all singular behaviours that can occur. Together with the known classification in the regular case, this completes the solution of the problem posed by Sophus Lie in 1882.
Read the full abstract

Artificial Intelligence

Interchangeable in Role, Not in History: Replacing Members of Formed LLM Teams

Multiagent LLM systems often treat a role as a slot: when a member is replaced, any agent competent at the role is expected to fill it. We test this view on teams whose members keep short written notebooks. A one-seat swap replays a formed, frozen team with its own member, with a same-role newcomer carrying another team's notebook for that role, or with a newcomer carrying none. In cooperative cooking, wildfire response and medical question answering, the experienced newcomer, averaged over seats, outscores the one without a notebook in every setting, so the usual comparison would approve it. Against the original team, the same replacement falls short in every setting, and the team spends more messages per unit of progress. On the hardest cooking level, the shortfall appears only after the team has worked together, persists across model families and decoding temperatures, and narrows by two thirds with added planning-only members. Because the team's history is written down, we can take it apart. In the default configuration on that level, notebooks from the same donor teams leave a newcomer that leads, holding the recipe, worse off than no notebook and one that follows better off. The notebook lines about the partner reproduce this whole difference between the seats. Whether an agent can replace a teammate depends on the seat and the team it enters.
Read the full abstract

All authors Jianxin Gao, Runze Li, Tianyi Yu, Liangwei Ren, Bohan Chen, Zining Wang

Software Engineering

Ritgard: T(r)opical Islands of Socio-Technical Artifacts on GitHub

A software project is more than just code. Non-code artifacts often document the human processes and decisions behind source code. The rationale behind a library change, an architectural decision, a problem encountered by a user are all examples of information typically present in socio-technical artifacts (STAs), created and persisted in channels separate from the repository itself (yet sometimes very close—e.g., GitHub Issues with GitHub repositories). These STAs are a trove of information about the project's architecture and its evolution, containing details and insights that code alone cannot provide. Unfortunately, this information is not easily extracted and explored as STAs are frequently fragmented over different communication channels, and are written in natural language. We present Ritgard, a tool that mines GitHub repositories for their STAs, namely Issues, Pull Requests, and Discussions, and visualizes them as 3D islands covered with trees. Each tree represents a single artifact and each island is a topic extracted from the artifacts through a combination of text embedding and text summarization. The terrain of the islands rises out of the ocean as the topic becomes active and sinks back in when it becomes stale, thus depicting the evolution of features and concerns throughout the project's lifetime. We describe the tool's usage and implementation, showing the numerous technical challenges behind Ritgard's visualization.
Read the full abstract

All authors Adam Štěpánek, Marco Raglianti, Jan Byška, Barbora Kozlíková, Michele Lanza

Analysis of PDEs

Asymptotics for the $k$-Hessian Eigenvalue on the Unit Ball

We compute the limit of the eigenvalue of the $k$-Hessian operator on the unit ball in $\mathbb{R}^n$ when $k\rightarrow \infty$, assuming that the ratio $\frac{n}{k}$ remains fixed. When $n< 2ke$, we moreover identify the limit of the corresponding eigenfunctions. We also derive monotonicity results and consider when the ratio varies in $n$.
Read the full abstract

Disordered Systems and Neural Networks

Connecting heterogeneous dynamics with local entropy

Establishing a robust and physically interpretable link between static structure and heterogeneous relaxation dynamics remains a fundamental challenge in glass physics. Here, we introduce a weighted pair-entropy descriptor based on the conventional two-body excess entropy. For this, we multiply the integrand used to calculate the excess entropy by a weight function that is directly related to the length scale of the pair correlation function. This multiplication does not affect the contribution of the short range order to the local excess entropy, but allows to take into account the structure present on intermediate distances, i.e., the medium-range order. For a canonical two-dimensional Lennard-Jones glass former, the resulting descriptor exhibits a strong correlation with particle-level dynamical propensity at long times (multiples of the alpha-relaxation time), with a maximum structure–dynamics correlation reaching about 0.9, substantially outperforming the predictive power of the conventional local pair excess entropy. These results demonstrate that incorporating a physically motivated structural length scale into entropy-based descriptors markedly enhances their predictive power while preserving physical interpretability. Our findings provide a simple and general framework for investigating structure–dynamics correlations in glass-forming systems.
Read the full abstract

All authors Jun Wu, Walter Kob, Yujie Wang, Zhen Zhang

Artificial Intelligence

Don't Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference

Layer dropout (a.k.a. stochastic depth) has been shown to enable faster training, higher accuracy, and robustness to zero-shot layer pruning in both language and vision transformers. However, as models and datasets have scaled, dropout - particularly layer dropout - has largely disappeared from large language models (LLMs) pre-training recipes. While some prior work has reported that dropout can degrade accuracy, no comprehensive study has quantified, let alone mitigated, this effect. In this study, we show that layer dropout should be used in state-of-the-art LLM training, establishing best practices and scaling analysis for both training and post-training benefits. Concretely, with optimal layer distribution, time schedule, and optimizer hyperparameters, we observe that at the same training FLOPs layer dropout leads to lower loss. For a given number of training steps, LLMs can achieve lower or similar validation loss while saving upto 25% of training FLOPs. Moreover, layer dropout enables significant post-training optimizations, such as early exit, intermediate-layer skipping, and self-speculative decoding, yielding up to 1.5x inference speedup with negligible accuracy loss. Across more than 2400 training experiments, spanning models from 271M to 8.2B parameters and datasets up to 160B tokens, we demonstrate that these findings extend reliably to large-scale training regimes. All pre-training experiments were run on Cerebras CS-3 systems.
Read the full abstract

All authors Mostafa Elhoushi, Alex Pretko, Nolan Dey, Bin Claire Zhang, Gavia Gray, Gurpreet Gosal, Abdulrahman Mahmoud, Shane Bergsma, Joel Hestness