Peer Revenue Publish or profit

Papers

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

All of arXiv

Monday 7 September

Robotics

ICI-VLA: In-Context Imitation with Spatiotemporally Aligned Demonstrations for Vision-Language-Action Models

Vision-Language-Action (VLA) policies are commonly adapted to new manipulation settings through additional gradient updates, which limits rapid deployment when task-specific data or compute is scarce. We present ICI-VLA, a training and retrieval framework that equips a text-action VLM with few-shot test-time adaptation through in-context demonstrations. Unlike mainstream VLA designs based on action-specific multimodal fusion, ICI-VLA retains the native text-generation interface. ICI-VLA updates its parameters only during offline training; at inference, the policy remains fixed and conditions action generation on retrieved micro-demonstrations. The framework decomposes long trajectories into short, semantically labeled examples and trains an RD-Encoder with positives mined by Dynamic Time Warping (DTW), aligning the retrieved context with the phase and geometry of the current subtask. We further introduce Target Action Masking, a context-corruption objective designed to reduce direct action copying and increase reliance on the current observation. ICI-VLA reaches average success rates of 97.7% on LIBERO and 60.4% on RoboTwin 2.0, exceeding the highest reported baseline average on RoboTwin 2.0 by 19.3 percentage points. It also achieves 83.2% across four physical tasks. These results indicate that a fixed VLA policy can benefit from conditioning on spatiotemporally aligned demonstrations at test time.
Read the full abstract

All authors Songhua Yang, Ziyu Liu, Xuetao Li, Ruqi Xiao, Kangxin Zhu, Miao Li

Combinatorics

The Neighbor Graph of Linear Complementary Dual (LCD) Codes

Linear complementary dual (LCD) codes form an important class of linear codes with applications in cryptography, classical error correction, and quantum coding theory. In this paper, we study the neighbor relation on LCD codes over finite fields and the graph induced by this relation, where two codes are adjacent whenever they intersect in codimension one. We determine the number of neighbors of an LCD code that are also LCD, and we use this result to analyze the structure of the corresponding neighbor graph. In particular, we prove its regularity over arbitrary finite fields and establish further regularity properties for its main structural subgraphs in the binary and odd-characteristic cases. These results provide a graph-theoretic framework for the study of LCD codes and reveal a strong combinatorial regularity in their neighborhood structure.
Read the full abstract

All authors Javier de la Cruz, Anna-Lena Horlemann, Marc Newman, Carlos Vela Cabello, Wolfgang Willems

Quantum Physics

Geometry Dependence of Error Thresholds in Two-Dimensional Toric Codes

We investigate how the lattice geometry influences the error thresholds of two-dimensional toric codes. In the presence of bit- or phase-flip errors, the toric code maps onto the two-dimensional random-bond Ising model (RBIM). We determine the critical behaviour of the RBIM using replica-exchange Monte Carlo simulations. While previous studies have explored thresholds for general lattice geometries under suboptimal decoders such as minimum-weight perfect matching, the rapid development of near-optimal decoders makes resolving the ultimate, maximum-likelihood code capacity relevant. We compute these optimal error thresholds on the Nishimori line for the square, honeycomb, triangular, dice, and kagome lattices. Owing to lattice duality, the phase-flip threshold on a given lattice is equivalent to the bit-flip threshold on its dual. We find that the optimal thresholds of dual-lattice pairs display a characteristic duality-driven splitting around the self-dual square-lattice, mirroring the qualitative behaviour observed for suboptimal decoders. While the average coordination number has the strongest impact on the error threshold, our results demonstrate that the detailed arrangement of vertices and plaquettes also plays a significant role in determining its precise value.
Read the full abstract

All authors D. Lessing, A. Langheld, C. Krämer, J.A. Koziol, K.P. Schmidt

Mathematical Physics

Entropic repulsion to the middle layer

We consider $\nabla\varphi$ height functions with even and convex interaction energy $W$ on the lattice $\mathbb{Z}^d$, which are restricted to take values in the set $\{-S, \ldots, S\}$ for some integer $S \ge 1$. We study the effect of entropic repulsion, which tends to push the spin values to the middle layer. We prove that the model has a unique Gibbs measure, with exponential decay of correlations, in the cases: (a) Dimension $d=2$ at all temperatures. (b) Dimensions $d\ge3$ at all temperatures, for a wide class of $W$ with non-increasing second derivative, including the family $W(x)=|x|^p$ for $p\in[1,2]$. (c) Dimensions $d\ge 3$ at both low and high temperatures, $T\in (0,\frac{W(1)d}{4(\ln d+\ln 8)})\cup(d W(2S),\infty]$, with the normalization $W(0)=0$. At low temperatures, our proof provides an alternative to Pirogov–Sinai methods. Conversely, we exhibit a class of even and convex interaction energies $W$ which, in high dimensions and suitable temperature regimes, have multiple Gibbs measures. Though uniqueness may fail, we show that the magnetization of every Gibbs measure lies in $(-\frac{1}{2},\frac{1}{2})$. This implies the delocalization of the model restricted to take values in $\{0,1,\ldots\}$ (i.e., conditioned to lie above a floor) for all dimensions, any even and convex $W$, and all temperatures. Our methods extend to additional setups: We prove that height functions taking values in the real interval $[-1,1]$ always have a unique Gibbs measure, a result previously proved only for the quadratic interaction. For height functions taking values in $\{-S+\frac{1}{2}, \ldots, S-\frac{1}{2}\}$, $S \ge1 $ integer, we prove that the magnetization of every Gibbs measure lies in $(-\frac12, \frac12)$. The special case $W(x)=x^2$ of our results addresses questions left open in the work of Bricmont–El Mellouki–Fröhlich (1986).
Read the full abstract

Distributed Computing

GPU-Accelerated Hypergraph Partitioning and Placement to Map SNNs on Neuromorphic Hardware

SNNs running on neuromorphic hardware use spikes to achieve sparse and energy-efficient communication over a mesh of cores. In turn, system performance heavily depends on the assignment of neurons to cores: the mapping. Since hardware features inter-core multicast and intra-core replication of spikes, we model SNNs as hypergraphs to exploit both opportunities for reducing communication traffic. Mapping thus comprises two NP-hard problems: hypergraph partitioning and placement on the lattice of cores. High-quality solutions to both are critical, yet increasingly difficult as networks scale to millions of neurons. Therefore, we propose a GPU-accelerated pipeline for SNN mapping: a multi-level partitioning scheme is devised around hardware constraints, while placement is initialized through recursive bisection, followed by refinement pulling together strongly connected cores through repeated swaps. Model-based experiments show upwards of 16% lower latency and 42% lower energy for spike movements over existing sequential tools, while our parallel mapper is on average 18-280x faster.
Read the full abstract

General Relativity and Quantum Cosmology

Matter-induced global regularity in non-polynomial quasi-topological gravity with Born-Infeld electrodynamics

We construct exact static, spherically symmetric charged solutions in four-dimensional non-polynomial quasi-topological gravity coupled to Born-Infeld electrodynamics. We focus on the model $h(p)=p/(1+\ell^{2}p)$, whose vacuum branch develops a curvature singularity at a finite radius. We show that Born-Infeld nonlinearities can remove this singularity within a finite region of parameter space, yielding globally regular geometries with an asymptotically flat exterior and a finite-curvature AdS-type core. The regular sector contains both horizonless configurations and RBHs, separated by a degenerate-horizon boundary. We further identify a continuous branch of regular black holes with a triple-degenerate inner horizon and a simple outer event horizon, satisfying $\kappa_-=0$ and $\kappa_+\neq0$. These results provide a converse example to cases in which introducing charge spoils the regularity of a black hole that is regular in vacuum. In the present model, Born-Infeld electrodynamics instead removes the finite-radius singularity of a gravitational branch that is singular in vacuum and supports globally regular charged geometries, including regular black holes with nontrivial inner-horizon structure.
Read the full abstract

All authors Hong-Lin Liu, Zhong-Wen Feng, Qing-Quan Jiang, Xia Zhou, Xue-Ling Mu

Machine Learning

Efficient Exploration Is Enough

This work introduces an alternative view of efficient exploration and studies its theoretical and empirical implications in the absence of extrinsic rewards. Specifically, we define efficient explorers as agents that prioritize generating generalizable experience, i.e., data that supports learning models capable of predicting and adapting across the environment. This allows us to analyze efficient exploration through the lens of prediction and generalization. Theoretically, we demonstrate that optimally efficient explorers naturally schedule their trajectories to visit the most informative and learnable regions first. Empirically, we show that optimizing for these agents gives rise to an automatic curriculum of progressively more complex behaviors, even in relatively simple environments. These results indicate that pursuing this purely intrinsic objective alone is enough to drive the emergence of highly sophisticated behaviors. We believe that this new framework provides a principled mechanism by which agent-environment systems may sustain an open-ended process of increasingly complex behavior without external rewards, tasks, or objectives.
Read the full abstract

All authors Mikel Malagón, Jon Vadillo, Josu Ceberio, Michael Bowling, Jose A. Lozano

Instrumentation and Methods for Astrophysics

Optical Flashes from Beam-Driven Light Sails with the Roman, Rubin and Euclid Observatories

The primary challenge of rocket propulsion is the burden of accelerating the spacecraft's own fuel. Light sails leave the propellant at home, with the achievable speed set by the sail area, the thermal tolerance of its material, and the power of the driving array. In GL15 we showed that leakage from a microwave array driving such a sail between habitable worlds produces Jansky-level radio transients lasting tens of seconds at $100$~pc. We take that leak to optical and near-infrared, where Fresnel matching would shrink the aperture to $\sim\!20$–$100$~m at $1\,\mu{\rm m}$, but the intensity on that aperture is $2\times10^{8}\,{\rm W\,m^{-2}}$, five orders above published directed-energy loadings. Spreading $1.5$~TW at $\sim20\,{\rm kW\,m^{-2}}$ is a $\sim10$~km phased array, pushing the emitters; Lubin's $10^{3}\,{\rm W\,m^{-2}}$ loading is a $40$~km array, comfortable but more expensive. A tenth-wave delay on $40$~m tiles sized so one tile still covers the sail leaks $0.054$ of the power ($80$~GW) into a halo. The typical Galactic detection is a single $\sim30$~s peak at $m_{\rm AB}\simeq19.4$ in F146 at $8$~kpc; a pair $87$~s apart, about one in five, is the confirmation test. Roman's Galactic Bulge Time Domain Survey reaches $109$~kpc at $8\sigma$ on a faint host, so a pointed flash of that halo is visible from anywhere in the Galaxy if the beam points at us. For $10$–$40$~km optical beamers held to $\sim\lambda/10$ at $v_{\rm max}\gtrsim250\,{\rm km\,s^{-1}}$, $N_{\rm det}=1$ wants $\Gamma_{6{\rm h}}\simeq16$ and a null search limits $\Gamma_{6{\rm h}}\lesssim49$, a factor-of-three window in launch rate. The real threshold is set by PSF-coincident stellar and instrumental transients. If intensity-limited optical beamers are commonly employed in our galaxy, this activity could be revealed by Roman, Rubin and Euclid at no additional observing cost.
Read the full abstract

Artificial Intelligence

From Simulated Citizens to Simulated Deliberation: Challenges in Representation and Interaction

Can LLM agents simulate how a real public would deliberate? For such simulations to be meaningful, agents must reflect the population's opinions, and their views must be shaped by what others say during discussion. We evaluate these requirements across six LLMs using national surveys, controlled deliberation experiments, and a real deliberative poll. Comparing census-grounded Korean personas against national surveys, we find that agents often produce overly concentrated opinions and fail to reliably reproduce demographic differences. During simulated deliberation, agents exchange arguments, respond to one another, and frequently change positions. Yet similar changes occur even when agents cannot see one another's contributions, while simply restating their starting positions sharply reduces movement. We further test whether these limitations persist in a real-world setting, using agents constructed from actual respondents in the 2017 Shin-Kori deliberative poll. Even with real respondent profiles, agents fail to reproduce the panel's initial opinion distribution or how its opinions change across survey waves. Our findings show that generating deliberative discussions, representing public opinion, and reproducing interaction-driven opinion change are distinct capabilities. These results suggest a different potential role for LLM deliberation: surfacing arguments and perspectives, rather than simulating a real public.
Read the full abstract

All authors Chaemin Jang, Junsik Min, Jaewoo Choi, Donggyu Lee, Haiin Lee, Yoonseo Choi, Junyoung Park, Namhee Kim, Hyunwoo Kim, Juho Kim, Nuri Kim, Jihee Kim

Soft Condensed Matter

Aggressive Phase Separation in Dense Mixtures of Passive and Active Particles

Using Vicsek-like self-propulsion rule, we study kinetics of liquid-liquid phase separation in mixtures of passive and active particles. For evolutions following temperature quenches of homogeneous configurations to the immiscible region of the phase diagram, we identify a remarkably strong dependence of the domain growth exponent on the choice of final state point. The singular dependence is indicative of the possibility of even an exponentially fast growth for suitable choices of system parameters. This striking observation, supported by finite-size scaling and other advanced analyses, is despite the fact that the overall mixture density is quite high that risks congestion with the prospect of slowing down particle transport. From flocking in pure biological systems to robotic swarming, these results are of much practical relevance. Theoretical pictures suitable for interpreting such high growth rates are discussed. In this regard, we discuss coarsening in the velocity field as well that describes how the density field coarsening follows the latter.
Read the full abstract

Computer Vision

SphereSOD: Geometry-Structure Coupled Learning for 360 Salient Object Detection

360{\deg} salient object detection (SOD) aims to accurately segment salient regions across a full field of view. However, equirectangular projection (ERP) introduces severe spatial distortion when mapping the spherical domain onto a planar representation. Existing methods mainly focus on compensating projection distortion while overlooking the interaction between panoramic geometry and salient object structure during feature perception and prediction refinement. To this end, we propose SphereSOD, an ERP-native framework that couples panoramic geometry with evolving salient structures. Specifically, spherical geometry governs feature sampling and spatial weighting, while coarse-grained saliency and contour prediction influence context aggregation during the progressive decoding process. SphereSOD first initializes deformable sampling based on spherical projection geometry and then employs bounded, content-adaptive offsets, yielding features that are better aligned with the underlying panoramic geometry. Subsequently, the decoder performs structure-guided context aggregation and progressive refinement to recover complete salient regions and accurate boundaries. Extensive experiments on three public 360{\deg} SOD benchmarks demonstrate state-of-the-art performance and a favorable accuracy-efficiency trade-off, supporting structurepreserving inference directly in ERP space as a promising alternative to projection-heavy panoramic pipelines.
Read the full abstract

All authors Junsong Zhang, Zhijie Shen, Shuai Zheng, Feng Li, Runmin Cong, Yao Zhao, Chunyu Lin

Combinatorics

Partial Petrial Polynomials of Bouquets

Gross, Mansour, and Tucker [European J. Combin., 95 (2021): 103329] introduced the partial Petrial polynomial of a ribbon graph $G$, denoted by $^{\partial}{\varepsilon^{\times}_G}(z)$. For a prime bouquet $B_n$, Yan and Li [Discrete Appl. Math., 375 (2025): 281-289] determined $^{\partial}{\varepsilon^{\times}_{B_n}}(z)$ when the intersection graph $I(B_n)$ is either the complete graph or a path, and provided an equivalent condition under which the lowest degree of the nonzero coefficient in $^{\partial}{\varepsilon^{\times}_{B_n}}(z)$ is $1$. In this paper, we determine $^{\partial}{\varepsilon^{\times}_{B_n}}(z)$ when the intersection graph is a cycle. Moreover, we present a complete characterization of the prime bouquets whose lowest nonzero term in $^{\partial}{\varepsilon^{\times}_{B_n}}(z)$ is of degree $2$ and determine the partial Petrial polynomial for the prime bouquets. As corollaries, we determine $^{\partial}{\varepsilon^{\times}_{B_n}}(z)$ when $I(B_n)$ is the complete bipartite and tripartite graph.
Read the full abstract

All authors Xiaoxiang Yu, Rong-Xia Hao, Jianbing Liu, Zhiguo Li

Rings and Algebras

The Pozhidaev and Cantarini–Kac Constructions of Simple $n$-Lie Algebras: Distinctions and Realizations

Let $n\geq 3$, let $H\subseteq\mathbb{C}^n$ be any additive subgroup spanning $\mathbb{C}^n$, and let $0\ne t\in H$. We study Pozhidaev's central simple $n$-Lie algebra $P(H,t)=\widetilde{\mathcal A}(H,t)/\mathbb{C}e_0$ without a finite generation or discreteness assumption on $H$. Its inner derivation algebra is the simple generalized divergence-free Lie algebra $\mathcal S(0,0,n;t,H)$. We prove that its space of inner-equivariant symmetric products vanishes and that every $1/n$-derivation is a scalar multiple of the identity. Using these invariants, we show that $P(H,t)$ is not isomorphic to any simple nonabelian $n$-Lie algebra defined on the underlying spaces of the $S$, $W$, or $SW$ constructions recorded by Cantarini and Kac. We also show that $P(H,t)$ is the quotient by the constants of the derived algebra of an explicit $S$-algebra on $\mathbb{C}[H]$. Finally, we realize Pozhidaev's second construction $E(H)$ over $\mathbb{C}$ as a $W$-algebra.
Read the full abstract

Computation and Language

We're Cooked! - Probing LLM Political Alignment Via Conflict-Framed Recipe Translation

Large language models (LLMs) are increasingly deployed for translation tasks, yet their implicit political positioning in such contexts remains understudied. We ask whether a single politically charged framing term, such as aggressor, enemy, neighbour, or coloniser is sufficient to trigger implicit political alignment in an otherwise apolitical task. We present a fully crossed factorial study in which eight models spanning Western, Chinese, and European origins are prompted to translate culturally attributed recipes into a target language left deliberately unspecified. Across 17 languages, four framing conditions, eight models, and 15,680 responses, we find that models do not simply decline or ask for clarification but resolve the ambiguity. Language resolution and reasoning behavior cluster meaningfully along model families: Western models hedge and deflect with vague justifications, Chinese models resolve conflicts silently, and Mistral Large emerges as a distinct profile combining high compliance with conflict-grounded reasoning. Sensitivity to framing terms is consistent across models: even subtle framing variation is sufficient to modulate behavior. Our findings urge caution when deploying LLMs for translation in conflict-adjacent contexts, where implicit political judgments may be made without any signal to the user.
Read the full abstract

Machine Learning

No-Regret Mixing of LRU and LFU with Optimal Switching Cost

Caching systems often rely on simple eviction policies such as Least Recently Used (LRU) and Least Frequently Used (LFU), which perform well in complementary request regimes. Recent policies such as LeCar and Cacheus combine LRU and LFU using ideas from the experts problem in online learning. Specifically, upon a miss, they randomize between the two eviction rules using probabilities derived from scores updated by tracking the history of past evictions. While these policies exhibit strong empirical performance, it remains unclear whether they are guaranteed, on every request sequence, to perform asymptotically as well as the better of LRU and LFU, i.e., whether they achieve sublinear regret with respect to this benchmark. We first show that LeCar suffers linear regret against an oblivious adversary, even with unbounded history. We then propose H-MC, a Hedge-based mixture of virtual LRU and LFU caches that preserves Hedge's selection probabilities, and hence its regret guarantees, while minimizing the switching cost among all joint selection rules with these marginals.
Read the full abstract

Strongly Correlated Electrons

Raman signatures of a non-reciprocal magnetic phase transition in Ca$_2$RuO$_4$

The magnetic behavior of Ca$_2$RuO$_4$ represents a unique puzzle due to the interplay of strong electronic correlations, magneto-elastic interactions, and large spin-orbit coupling. At low temperatures, an anomalous Mott insulating state emerges, characterized by a complex antiferromagnetic order with a collective amplitude excitation of the magnetic moment which has been discussed in analogy with the Higgs mode. We report here evidence of a magnetic first-order phase transition driven by an out-of-plane magnetic field along the crystallographic $c$-axis of Ca$_2$RuO$_4$. Raman measurements in magnetic field reveal the emergence of mode of magnetic origin and a concomitant modification of the coupling between a phonon and the Higgs amplitude mode. Both the Raman features are characterized by a non-reciprocal hysteresis in magnetic field. Surprisingly, the observed phase transition does not affect the Raman scattering from the in-plane magnons, indicating that the dipolar antiferromagnetic order is preserved. This is consistent with the onset of a new quadrupolar field-controlled state, whose fluctuations structure can trigger an hybridization between the lattice and the magnetic Higgs mode resulting in the observed Raman features.
Read the full abstract

All authors Giacomo Jarc, Giovanni Tartaglia, Francesco Gabriele, Filomena Forte, Anita Guarino, Angela Montanaro, Enrico Maria Rigoni, Nitesh Khatiwada, Costanza Lincetto, Gabriele Bartolini, Antonio Mastropasqua, Shahla Yasmin Mathengattil, Marco Malvestuto, Muhammad Waqee Ur Rehman, Rosalba Fittipaldi, Joachim Deisenhofer, Alexander A. Tsirlin, Antonio Vecchione, Mario Cuoco, Daniele Fausti

Metric Geometry

De-Höldering factorization

We study a factorization notion for Lipschitz functions between metric spaces in which such a function is written as a composition of a Hölder function and a function that suitably “undoes” the Hölder regularity. We show simple ways to construct such “de-Höldering” factorizations. If the identity mapping on a metric space $\mathcal{M}$ admits a de-Höldering factorization through a metric space $\mathcal{Z}$ that has a conical geodesic bicombing, then the class of metric spaces from which one can extend $\mathcal{Z}$-valued Lipschitz functions is shown to be contained in the corresponding class of metric spaces for $\mathcal{M}$-valued Lipschitz functions. As a quick consequence of these abstract permanence properties, we deduce that every $L_1$-valued Lipschitz function from a subset of $\ell_2$ can be extended to a Lipschitz function that takes values in $L_1$ and is defined on all of $\ell_2$, answering a 1992 question of Ball. By work of Makarychev and Makarychev, this implies that every weighted graph has a vertex cut sparsifier of size $n$ and quality $O(\sqrt{\log n})$, improving Moitra's 2009 bound. We also show that for every metric space $\mathcal{Z}$ that has a conical geodesic bicombing, any metric transform of a metric space $\mathcal{M}$ has $\mathcal{Z}$-valued Lipschitz extension modulus at most a universal constant multiple of the $\mathcal{Z}$-valued Lipschitz extension modulus of $\mathcal{M}$ itself, improving the 2002 bound of Brudnyi and Shvartsman.
Read the full abstract

Information Theory

Resource-Constrained Semantic-Aware Remote Estimation with Overlapping Sensor Coverage

We study semantic-aware remote estimation of multiple finite-state Markov sources observed by K sensors with overlapping coverage. The sensors share a time-division multiple-access uplink and differ in transmission reliability, delivery delay, and transmission budget. In each slot, the scheduler jointly selects a source and one of its monitoring sensors, or remains idle, to minimize the long-run average cost of actuation error subject to global and per-sensor transmission-frequency constraints. We formulate this problem as a finite average-cost constrained Markov decision process. We show that the transmission resource functions have rank at most K, although the global constraint may still restrict the feasible region. Consequently, the Lagrangian depends only on K effective transmission costs, and an optimal constrained solution can be represented using at most K+1 deterministic policy-recurrent-class components. We further characterize the piecewise-affine concave Lagrangian value function and derive projected dual subgradient ascent over an explicitly bounded multiplier set. Numerical results illustrate the value-function structure, the need for policy randomization in a representative instance, and the interaction between global and per-sensor transmission budgets.
Read the full abstract

Algebraic Geometry

Group action of Hochschild-Serre algebra and categorical reconstruction

We study the action of the Serre functor of smooth proper dg categories at their Hochschild-Serre algebra, and the invariant sub-aglebra of the Serre functor. As applications, we prove some theorems of categorical Torelli. Namely, let $\Ku(\X)$ be the Kuznetsov component of degree $d$ smooth hypersurface in weighted projective space $\mathbb{P}(a_0, a_1, \cdots, a_n)$, where the common maximal divisor $\gcd(d, \sum^{n}_{i=0}a_{i})=1$. We show the categorical Torelli for $\Ku(\X)$. We show that the $\mathbb{C}^{\ast}$ equivariant matrix factorization category associated with a quasi-homogeneous polynomial function $f$ that has an isolated singularity together with a twisted functor $\{1\}$ reconstructs $f$ up to an isomorphism.
Read the full abstract

Information Retrieval

EigenLI: Spectral Approximations to Late Interaction

Late-interaction models such as ColBERT achieve strong effectiveness by representing each document with many token-level vectors, but this expressivity leads to large indexing cost, storage footprints and expensive MaxSim scoring. We show that late-interaction representations exhibit an intrinsic low-rank structure: document token embeddings concentrate in a low-dimensional subspace that preserves most of the retrieval signal. Leveraging this observation, we introduce EigenLI, a spectral approximation framework that compresses late-interaction representations via document-specific low-dimensional subspaces. Unlike clustering or pooling methods, EigenLI identifies the dominant eigendirections of each document and uses them to construct reduced interaction representations. Empirically, $k$-EigenLI with $k \le 32$ outperforms k-means and Ward clustering based pooling methods on ColBERTv2 and AnswerAI-ColBERT-small; GTE-ModernColBERT exhibits a different tradeoff at $k=32$, where clustering methods perform better. The same spectral construction also yields EigenLI-SV, an ANN-compatible single-vector representation derived from the second-order summary of the reduced structure. Across multiple datasets and all three text models, EigenLI-SV consistently outperforms comparable single-vector surrogates such as MUVERA.
Read the full abstract

All authors Archish S, Sabyasachi Basu, Ankit Garg, Ravishankar Krishnaswamy, Kirankumar Shiragur

Human-Computer Interaction

From Human Factors to Human-Technology Factors: An HCI Perspective on Technology in Avalanche Safety

The avalanche safety literature has identified human factors that contribute to accidents, yet researchers note a persistent gap between these insights and effective product design. Meanwhile, interactive technologies are shaping backcountry decision making with or without grounding in theory and research. We broaden the discussion of human factors into “human-technology factors," examining how technology can both support and undermine judgment in avalanche terrain, from a human-computer interaction (HCI) perspective. We summarize relevant HCI concepts along four dimensions - attention, cognition, trust, and social interaction - and use them to revisit McCammon's FACETS framework, cataloging ways in which specific technologies may mitigate or exacerbate classic heuristic traps, grounded in accident reports and literature where possible. We identify recurring patterns, including technologies with dual-sided effects and a pervasive “digital expert halo."
Read the full abstract

Artificial Intelligence

Scoring Without the Engine: Validating a Deterministic, Manipulation-Resistant Content Score for Generative Engines, End to End

How do you validate a cheap, deterministic proxy for an oracle that is expensive, rate-limited, and non-stationary? We present a protocol built on adversarial falsification gates (negative control, dose response, bounded amplification, duplication penalty, length neutrality) that define and select the proxy, fitted on a training split and confirmed held-out; around them it bounds what the proxy can never resolve, and re-measures external causal evidence on the current oracle rather than assuming it. We demonstrate it end to end on Generative Engine Optimization, where the proxy is a deterministic content score, and one step fails on that domain exactly as the protocol is built to detect: re-measuring the only published causal anchors (2023 effect sizes) on ten modern engine families shows their levers move citation on none, so the anchors are an expired external check; recalibrating to the near-zero modern vector strips the score of its lever-responsive components. What survives is the gate-enforced response surface. The gates buy a measured property: on a 500-source benchmark of adversarial edits, amplifying the score's calibrated levers gains an attacker at most 6 points, and decreases with dose; single-lever amplification is provably bounded, while the cap and cross-lever sub-additivity are empirical findings consistent with it. On detection, web-spam baselines dominate and out-of-distribution attacks evade the score, so the deployable filter layers it over them. A query-conditioned skyline bounds the score's citation signal (within-query Spearman 0.11), repositioning query-agnostic scores as quality filters rather than citation predictors. A query-leakage bug in our first ranking evaluation and a failed confidence flag are disclosed and corrected; every number reproduces offline from released artifacts at zero marginal API cost.
Read the full abstract

Geophysics

SeisBench DAS: A machine learning framework for Distributed Acoustic Sensing

Fibre optic sensing, such as distributed acoustic sensing (DAS), has become a widespread technology for geophysical studies. To process the large-scale datasets produced by DAS, several machine learning methods have been proposed. However, without standardization of data and models, these methods lack comparability and interoperability. This introduces a gap between model developers and practitioners analyzing DAS data and inhibits adoption of deep learning for DAS. To address these limitations, here we present SeisBench DAS, an extension to the SeisBench library for machine learning in seismology. SeisBench DAS defines standard formats for DAS benchmark datasets, including standardised metadata and labels, and DAS models. It builds on the xdas framework for data ingestion and virtual array handling, and on PyTorch for reading and applying the machine learning models. Importantly, SeisBench provides an engine to efficiently apply deep learning models to diverse formats of DAS data, bridging the gap between model developers and practitioners. SeisBench DAS is designed as an open and extensible framework, allowing to easily incorporate future developments in deep learning for DAS.
Read the full abstract

Computer Vision

Heat Kernel Textures: the Geodesic Gaussians That Do Not Splat

3D Gaussian Splatting has recently revolutionised novel view synthesis as well as many other 3D vision methods and applications. Drawing inspiration from this representation, we now rethink textures to overcome the main issues of UV mapping while considerably lowering their memory footprint. Heat Kernel Textures (HKTex) eliminate UV unwrapping as well as their persistent issues of wasted UV space, seams, distortions, vertex-duplication, and varying resolution. Grounded in discrete Riemannian geometry and intrinsically defined on any manifold surface discretised as a triangular mesh, HKTex uses anisotropic heat kernels as geodesic equivalents to Gaussians. Like our kernels, also the optimisation of their position and the adaptive densification strategies were redefined to operate on the surface of the object to be textureised. Our novel representation is also fully integrated with a physically based renderer and can be optimised either from existing textures or multi-view images. Our project page and code are available at circle-group.github.io/research/HeatKernelTextures.
Read the full abstract

All authors Simone Foti, Caner Korkmaz, Stefanos Zafeiriou, Tolga Birdal

Nuclear Experiment

Independent Validation of Octupole Collectivity in radium-224 through lifetime measurements of low-lying negative-parity states

The nucleus $^{224}$Ra is a key benchmark for octupole deformation and for theoretical descriptions of enhanced Schiff moments in reflection-asymmetric nuclei. While Coulomb-excitation measurements have established strong octupole collectivity in $^{224}$Ra, theoretical models predict that its intrinsic electric-dipole moment should be strongly quenched by a cancellation between macroscopic and microscopic contributions. Direct fast-timing measurements of the low-lying $J^\pi = 1^-_1$ and $3^-_1$ states populated following the $\beta$-decay of $^{224}$Fr at TRIUMF-ISAC were performed. Using the LaBr$_3$(Ce) detectors of the GRIFFIN array, mean lifetimes of $\tau(1^-_1) = 444(6)$~ps and $\tau(3^-_1) = 460(18)$~ps were obtained. The corresponding reduced transition probabilities agree with values inferred from Coulomb excitation, but are determined with substantially improved precision. These results provide an independent validation of the electromagnetic matrix elements associated with octupole collectivity in $^{224}$Ra and confirm a strongly-quenched intrinsic dipole moment of $D_0 \simeq 0.032~e\mathrm{fm}$. The present measurements therefore provide a stringent experimental benchmark for nuclear-structure models used in the interpretation of Schiff moments and future searches for non-zero electric dipole moments.
Read the full abstract

All authors D. White, D. O'Donnell, A. Avaa, J. R. Murias, S. Murillo Morales, R. Umashankar, V. Vedia, C. Andreoiu, D. W. Annen, A. D. Ayangeakaa, G. C. Ball, V. Bildstein, M. Bowry, I. Dillmann, E. G. Fuakye, L. P. Gaffney, A. B. Garnsworthy, P. E. Garret, E. D. Geerlof, S. Georges, A. L. Grimes, G. F. Grinyer, G. Hackman, P. M. Jones, J. Liu, B. Olaizola, S. D. Olorunfunmi, F. Rowntree, N. K. Syeda, D. Shah, P. Spagnoletti, C. E. Svensson, F. Wu

Functional Analysis

Sparsity-preserving means on a homogeneous convex cone

We study two means on the five-dimensional Vinberg cone $\Omega$, viewed as a sparse cone of positive definite $3\times3$ matrices. The Cholesky-Vinberg mean is obtained from the simply transitive solvable group $H$ associated with $\Omega$. It is $H$-equivariant, and its interpolation curves are geodesics for two flat metric connections with torsion. One of the corresponding metrics is the canonical Hessian metric of the cone. The Log-Vinberg mean is defined by affine interpolation in the global logarithmic coordinates of the associated Vinberg algebra, or clan. The resulting Riemannian metric is complete and flat, and its weighted Fréchet means have explicit formulas. Both constructions preserve the zero pattern defining $\Omega$.
Read the full abstract

Game Theory

Finding Representative and Approximately Efficient Committees

In approval-based committee voting, proportional approval voting (PAV) is a well-studied rule that combines proportional representation with Pareto efficiency. However, computing a PAV committee is NP-hard, raising a natural question: Can the proportionality and efficiency properties of PAV be achieved via computationally efficient procedures? We make two contributions toward answering this question. First, building on the known proportionality guarantees of the local-search-based variant of PAV (or local PAV), we systematically study its efficiency properties. We show that local PAV committees are weakly Pareto optimal, meaning that no other committee is strictly preferred by every voter. We also identify limitations: Local PAV guarantees only a $2$-approximation to fractional Pareto optimality ($2$-fPO) and a $2/3$-approximation to the optimal PAV score, and both bounds are tight. In contrast, global PAV is Pareto optimal and satisfies the stronger $\alpha^\star$-fPO guarantee, where $\alpha^\star \approx 1.346$ is the unique solution of $\int_0^{\alpha^\star} \frac{1-e^{-y}}{y} \, dy = 1$, and this approximation is tight. Second, we design a polynomial-time algorithm that combines the best of these guarantees. The committee returned by our algorithm satisfies EJR$+$ (a proportionality guarantee), $\alpha^\star$-fPO, and weak Pareto optimality. It also achieves a $0.79$-approximation to the optimal PAV score, matching the best possible polynomial-time approximation assuming $P \neq NP$. Our algorithm works by pipage rounding a concave relaxation of the PAV objective and using that committee to initialize local PAV, thereby combining global approximation guarantees with local search stability.
Read the full abstract

Analysis of PDEs

The equivalence of solutions to fractional and logarithmic Helmholtz equations

We show the following equivalence of solutions to general fractional Helmholtz equations: $u$ is a solution to $$A^su=\lambda^su$$ for some fixed $0<s<2$ and some $\lambda>0$ if and only if $u$ is a solution to the same equation for all $s\neq0$. Furthermore, we show as well that $u$ equivalently solves the logarithmic Helmholtz equation $$\log(A)u=(\log\lambda)u.$$ Here, $A$ is a nonnegative, linear operator on a Banach space $X$. In particular, our results hold whenever $A=-L$, where $L$ is the infinitesimal generator of a $C_0$-semigroup in a Banach space $X$ satisfying mild assumptions that are typical in applications. As a particular case, we recover known results for the fractional Laplacian $(-\Delta)^s$ in $\mathbb{R}^n$. More importantly, we provide a list of examples of other fractional power operators in settings where the Fourier transform is not available and for which our theorems apply. These include fractional powers and logarithms of second order elliptic operators in bounded domains, Laplace–Beltrami operators on Riemannian manifolds, space-time master equations, the discrete Laplacian, and fractional derivatives.
Read the full abstract

General Topology

Semiflows deforming automorphisms groups

In this short note, we prove that, for any pair of finite groups $G$ and $H$, there exist a finite $T_0$-space $X$ and a semiflow $\varphi\colon [0,\infty)\times X\to X$ such that $\operatorname{Aut}(X)\cong G$, whereas $\operatorname{Aut}(\varphi_t(X))\cong H$ for every $t>0$. Thus, the symmetry group of a finite $T_0$-space and those of all its positive-time images can be prescribed independently.
Read the full abstract