Computer Vision
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 1 September
An image may be worth a thousand words, but most captioning models describe it in only a few. Modern vision-language models produce fluent high-level captions, yet routinely miss the attributes, counts, textures, materials, and spatial relations that make an image visually specific. Recent multi-stage systems recover some of these details through generation, decomposition, verification, and rewriting, but they do so at the expense of substantially higher inference latency. We propose SimLoss, a reference-free embedding-space objective for single-pass fine-grained image captioning. SimLoss trains a vision-language model to align its projected hidden-state representation with a frozen image embedding through an InfoNCE contrastive loss, supplying a dense visual supervision signal before any text is decoded, and requiring neither human-written fine-grained captions nor pseudo-captions from a multi-stage pipeline. We instantiate it as SimLoss FFT, which backpropagates through a locally available embedding model, and SimLoss GRPO, which treats that model as a black-box reward. Compared with single-pass, multi-stage verification, reward-optimized, and perception-aware baselines, the fully differentiable fine-tuning variant, SimLoss FFT, achieves the highest precision while nearly matching the F1 score of the multi-stage method, all while retaining single-pass inference and running roughly 20 times faster than the multi-stage pipeline. The reward-based variant SimLoss GRPO attains the strongest recall. Together, these results show that embedding-space supervision can recover the quality of multi-stage verification at the latency of a single-pass captioner.
Machine Learning
CRAFT: Fine-Tuning Pre-hoc Explainability in AI-native 6G RAN
The next generation of mobile networks is envisioned as fully AI-native, with AI-RAN architectures embedding small language models (SLMs) to perform reasoning over real-time telemetry. The state-of-the-art training paradigms for telecom LLMs, exemplified by RANSTRUCT-style supervised fine-tuning (SFT) on curated instruction data, are limited to post hoc rationalization. Here, the explanations, when produced at all, are generated after or independently of the decision, leaving the decision process unauditable. Pre-hoc reasoning, where a causal reasoning trace is produced before the output label, is preferable, and the broader LLM reasoning literature has made real progress toward it via RL methods such as Group Relative Policy Optimization (GRPO). Here we observe that transplanting this recipe into the telecom setting runs into a cold-start barrier: SLMs either learn to output the desired format or learn to predict the label, but rarely both. We identify this barrier and propose CRAFT, which stands for Cold-start Reasoning Alignment via Fine-Tuning, a data-centric method to autonomously generate a verified dataset of (input, trace, label) triplets. CRAFT fine-tunes SLMs on this verified data using low-rank adaptation (LoRA), requiring substantially less compute and wall-clock time than GRPO-based methods. On the TRACTOR and IC xApp telecom datasets, CRAFT achieves up to 86.5% and 94.6% for accuracy and F1 with no parse failures, while direct GRPO and SFT+GRPO fail to exceed 28% and 53.5% F1 with multiple parse failures. We further show that CRAFT-initialized policies serve as a robust foundation for subsequent GRPO fine-tuning, as under diverse reward functions the performance remains consistent with no parse failures. Finally, we demonstrate that CRAFT consumes 59% less energy than GRPO-based baselines, making it a sustainable path to deployable, auditable AI in 6G RAN.
Read the full abstract
General Relativity and Quantum Cosmology
Directional Response Optimization through Linear Recombination of Time-Delay Interferometry Channels in Space-based Gravitational Wave Detection
Space-based gravitational-wave detectors such as LISA, Taiji, and TianQin employ time-delay interferometry (TDI) to cancel laser-frequency noise for unequal-arm constellations. Since different TDI observables exhibit distinct sky responses, a linear combination of candidate channels can enhance the average response over one sky region while suppressing that of another. We construct a frequency-domain response matrix for TDI combinations, average it across target and suppressed sky regions, and derive the optimal weights via a generalized eigenvalue problem that maximizes the ratio between these two regional responses. At millihertz frequencies, examples with the $A$, $E$, and $T$ channels, the Sagnac combinations $\alpha$, $\beta$, and $\gamma$, and 16-links TDI show that a sky-region null and a large regional contrast are possible near the chosen frequency, with eigenvalues $\rho$ ranging from $\mathcal{O}(10)$ for small bases to $\mathcal{O}(10^2)$ for the larger set. The method is therefore expected to be well suited to nearly monochromatic sources such as the resolved Galactic double white dwarf binaries in the millihertz band. The Target-to-Suppression Ratio (TSR) peaks near the design frequency and falls quickly away from it, so the optimized weight vector is inherently narrowband and suited to targeted searches around a chosen frequency and sky direction.
Read the full abstract
Computation and Language
Quit While You're Ahead: Quit for Efficient Candidate Generation in Machine Translation Reranking
Reranking methods, such as Minimum Bayes Risk (MBR) decoding and Quality Estimation (QE) reranking, have been widely used in modern neural machine translation (NMT) to select an output from a set of candidate hypotheses. However, the performance gains come at the cost of high inference latency. Existing acceleration methods target MBR decoding and reduce only the reranking computation, leaving QE reranking unaddressed and candidate generation—which can be the larger computational bottleneck—largely untouched. In this work, we propose Quit (Quantifying Uncertainty for Incremental Termination), a novel early-stopping strategy for the entire generation–reranking pipeline. Quit treats candidate generation as a sequential decision-making process under uncertainty. It incrementally generates and reranks candidates, stopping when the best reranking score stabilizes. Comprehensive experiments with three NMT models across 19 language pairs show that Quit achieves end-to-end speedups of $1.47$–$2.66\times$ for MBR decoding and $3.43$–$4.12\times$ for QE reranking while preserving automatic metric scores.
Read the full abstract
Soft Condensed Matter
Bond-number-controlled durability of cohesive granular materials under repeated vibration
Cohesive granular materials derive their mechanical stability not only from the strength of individual interparticle bonds but also from the number of bonds forming the load-bearing network. However, these two effects are difficult to separate experimentally because conventional control parameters, such as liquid content, generally alter both simultaneously. Here, we use a mixed granular system composed of cohesive and noncohesive grains to control the cohesive bond number while keeping the bond strength approximately unchanged. We investigate the failure lifetime under repeated vibration and find that the number of cycles to failure, $N_f$, depends strongly on the mixing ratio $\alpha$. In a mean-field picture of random mixing, the fraction of cohesive contacts scales as $\alpha^2$, and $N_f$ increases approximately exponentially with $\alpha^2$. By contrast, although the lifetime tends to decrease with increasing vibration intensity $G$, its dependence on $G$ is comparatively weak over the present experimental range. Remarkably, although the Young's modulus is nearly independent of $\alpha$ above the rigidity threshold, the lifetime continues to increase strongly with $\alpha$. This demonstrates that mechanical rigidity and durability against repeated perturbations exhibit distinct dependences on the cohesive network. These results identify bond number as a key control parameter for the durability of cohesive granular materials.
Read the full abstract
Superconductivity
Microscopic Origin of Pressure-Enhanced and Robust Superconductivity in Infinite-Layer La$_{0.8}$Sr$_{0.2}$NiO$_2$
Recent transport measurements on freestanding La$_{0.8}$Sr$_{0.2}$NiO$_2$ membranes revealed a broad superconducting dome extending from ambient pressure to 210 GPa, with an onset transition temperature reaching 74.5 K near 146 GPa. Using first-principles calculations, a pressure-dependent two-orbital model, and self-consistent FLEX calculations combined with the linearized Eliashberg equation, we determine how compression modifies the pairing tendency. Pressure increases the kinetic-energy scale, reduces $U_x/t_1$, strengthens interlayer hybridization, and transfers holes from the La/Sr-derived charge reservoir to the correlated Ni sector. Within the present low-energy description, the increasing kinetic scale and the approach to optimal intermediate coupling account for the initial enhancement of pairing, whereas pressure-induced self-doping into the overdoped regime is primarily responsible for its high-pressure suppression. Despite a pronounced three-dimensionalization of the Fermi surface, the pairing-relevant spin susceptibility remains weakly dependent on $q_z$ and peaked near $(\pi,\pi)$. Consequently, the Ni-$d_{x^2-y^2}$-dominated $d$-wave pairing state remains stable over the calculated pressure range. These results provide a unified microscopic interpretation of both the superconducting dome and its unusual robustness under megabar compression.
Read the full abstract
Signal Processing
Real-Time Neuromorphic Spectrum Intelligence Simulator
We present the Real-Time Neuromorphic Spectrum Intelligence Simulator (RT-NuSIS), a modular framework to study spiking neural network (SNN) and memristor-inspired agents for dynamic spectrum access under constrained energy budgets and adversarial conditions. RT-NuSIS couples leaky integrate-and-fire neuronal dynamics, memristive synaptic models, physics-informed energy-harvesting models (triboelectric and RF), and adversary models including jamming and Byzantine behavior. We formalize the simulator mathematically, prove boundedness, present a mean-field adversary threshold, analyze per-step complexity, and provide a reproducible benchmark harness for energy-per-inference, latency, and robustness metrics. The codebase is modular, deterministic by seed, and designed for large-scale event-driven simulations.
Read the full abstract
Artificial Intelligence
Socrates went Nuclear: Comparing Interaction Strategies for AI systems in a Learning Context using Brain Sensing
Does unrestricted AI access bypass the cognitive effort required for learning, or does it streamline knowledge acquisition? This paper reports on a study where we compare three designs for user-AI interaction in a learning context: (1) an unrestricted conversational bot like ChatGPT, (2) a pedagogically constrained bot that guides through hints without giving final answers, which we refer to as the Socratic mode; and (3) a non-conversational adaptive tutoring system that adjusts difficulty in real-time based on the user's cognitive engagement derived from the brain signals. Fifty study participants were tasked with learning about nuclear safety protocols, a domain chosen for its zero-prior knowledge baseline. The participants progressed through an instructional video, a pre-test, an AI-driven assessment phase, which varied in the three conditions, and an immediate post-test. The nature of the questions centered primarily on factual knowledge acquisition, but it still required participants to have a global understanding of the concepts in order to answer the questions correctly. A Muse headband was used to derive the cognitive engagement of all users in all conditions. The unrestricted chatbot produced higher learning gains (delta) than both constrained modes (p < .03, d > 0.80), while the adaptive condition generated significantly higher EEG engagement (p = .018). The cluster analysis of chatbot usage and discussion patterns by users showed that most participants in the unrestricted-mode adopted a direct answer-retrieval strategy, while participants in the Socratic-mode initially attempted to reason through the hints before progressively disengaging. Consequently, this also suggests that the success of the unrestricted AI is not an evidence of deeper learning, but rather a result of the immediate post-test evaluation after the training phase.
Read the full abstract
Combinatorics
A 21-Coloring of the Plane Without Monochromatic Unit-Area Rectangles
Erdős and Graham asked whether every finite coloring of the plane must contain a monochromatic rectangle of any prescribed area. Kovač gave a negative answer by constructing a $25$-coloring with no monochromatic rectangle of area $1$. We reduce the number of colors to $21$ by replacing the square cells in his construction with regular hexagons.
Read the full abstract
High Energy Physics (Phenomenology)
Spin density matrix for high-spin particles
The spin density matrix (SDM) characterizes the spin state of a particle, which provides important information for understanding its production mechanism. A general discussion of the SDM under parity conservation, and specific SDM parameterizations for spin-1 and spin-2 particles are provided, which respect all the expected constraints. The imaginary (real) part of SDM is related to the degree of odd-rank (even-rank) polarization, with specific expressions given for spin-1 and spin-2 cases. Examples of applications of the SDM to angular analyses of several decay processes with different spin-parity hypotheses are also provided, which could be helpful in experimental data analyses.
Read the full abstract
Computation and Language
Enoki: Efficient Multi-Level Hallucination Detection
Ensuring factuality remains a critical challenge for deploying LLMs in high-stakes settings. Existing hallucination detectors usually operate at a single level: claim-level methods provide interpretable factual units, while span-level methods localize unsupported text. Bridging these views is costly, as LLM-heavy pipelines require multiple decomposition and verification calls, and modular systems need additional claim-to-span alignment. We propose Enoki, an Open Information Extraction framework for multi-level hallucination detection. Enoki extracts text-anchored relational facts, verifies them against evidence, and projects unsupported facts back to hallucinated spans. This shared representation enables claim-level verification and span-level localization without requiring separate alignment. Enoki supports LLM-based, encoder-based, and rule-based extraction regimes, balancing accuracy and inference cost through a common interface. Experiments show that Enoki remains competitive with strong claim-level systems while using fewer resources and achieves superior performance on fine-grained span- and entity-level localization. We also release EnokiQA, a dual-granularity dataset with aligned claim-level verification and span-level localization annotations.
Read the full abstract
Materials Science
Lead-free piezoelectric perovskites for arterial-pulse e-skin: from configurational complexity to equivariant machine-learning potentials
Continuous, non-invasive monitoring of the arterial pulse is a clinical priority for cardiovascular disease, the leading cause of global mortality. Flexible piezoelectric electronic skins can transduce the 1-10 kPa pressure wave into a self-powered voltage, but the best-performing piezoceramics are lead-based, and their toxicity is incompatible with skin contact and with tightening RoHS/REACH regulation. Among lead-free alternatives, the BaTiO$3$-based solid solution BZT-BCT reaches $d{33} \approx 620$ pC/N near its tricritical morphotropic phase boundary, rivalling soft PZT while remaining biocompatible. Exploiting this in a wearable confronts a sensitivity-flexibility paradox and three computational walls: the combinatorial explosion of atomic configurations in a disordered solid solution, the band-gap error of affordable density-functional approximations, which corrupts leakage and insulation estimates, and the 0 K nature of standard calculations against a 310 K operating temperature. We review lead-free piezoelectrics, morphotropic-boundary physics and fabricated flexible devices, then argue that equivariant machine-learning interatomic potentials — coupled to a tiered functional hierarchy and finite-temperature lattice dynamics — can survey the full configurational ensemble at body temperature and close the gap to a clinically viable lead-free pulse sensor.
Read the full abstract
Programming Languages
Predicting Program Exit Code with LLMs and Programming Language Semantics
Large language models (LLMs) have shown proficiency in various software engineering tasks, such as code generation and translation. However, a key limitation in their performance may be their (lack of) understanding of programming-language semantics. Even when explicit semantics are given, it remains unclear whether LLMs apply those rules or lean on priors learned during pre-training instead. We study if LLMs lean on priors or given semantics with a novel task–Program Executability Prediction (PrEx)–that asks models to predict whether a program is semantically valid or invalid (and, if invalid, which formal rule it violates) given the program's syntax and operational semantics. Because PrEx requires both valid and invalid programs, we build a dataset with systematically generated invalid transformations derived from valid programs. We evaluate open-source coding LLMs under two semantic formalisms and two semantic shifts across Human-Written, LLM-Translated, and Fuzzer-Generated program splits. Our findings show that LLMs lean on pre-training priors rather than systematically applying the given rules, performing especially poorly on modified semantics and degrading further as program complexity increases. PrEx is available at https://github.com/EngineeringSoftware/prex.
Read the full abstract
Artificial Intelligence
Same Request, Different Boundary: Evaluating Cybersecurity Assistance across Conversational Contexts
Cybersecurity assistance is dual use, requiring decisions about when to answer, limit, or refuse a request. Prior work studies refusal evaluation and conversational jailbreaking, but how conversational histories shift assistance boundaries in both directions for identical cybersecurity requests remains underexplored. We introduce 3R-Bench, a benchmark of 150 real-world cybersecurity requests evaluated across eight LLMs under direct prompting and two predefined conversational settings. We separately code refusal, bounded assistance, compliance, and provider blocking to characterize delivered assistance. On 400 matched ambiguous request-model pairs, compliance is 75.3% directly, 57.8% after refused precedent, and 81.0% after accepted precedent. A smaller repeated-generation panel reproduces the refusal-acceptance gap. On a balanced 75-request panel reusing the direct baseline, compliance decreases from 74.2% directly to 57.2% after neutral task-decomposition dialogue and 60.3% after failure-ended dialogue. These findings show that cybersecurity assistance varies with conversational history and model deployment, motivating evaluation beyond direct-request refusal rates.
Read the full abstract
Machine Learning
GeoPAR: Large-Scale Multi-Agent Combinatorial Optimization with Geometry-Guided Parallel Autoregressive Learning
Multi-agent combinatorial optimization problems are notoriously challenging due to their NP-hard nature. Recent parallel autoregressive neural solvers improve inference efficiency by allowing agents to make decisions simultaneously, but their performance often degrades on large-scale instances. This is largely attributable to weak modeling of local geometric structures and the fact that conflicting task selections are handled only after action generation. To address these limitations, we propose GeoPAR, a geometry-guided parallel autoregressive reinforcement learning framework for scalable multi-agent combinatorial optimization. GeoPAR integrates three key components: (1) a projection-window sparse geometry mechanism that builds lightweight local candidate neighborhoods through multi-directional projections, (2) sparse edge-biased attention that injects these geometric relations into node representations, and (3) cache-guided conflict-aware assignment that reuses the geometric cache during decoding to suppress duplicate selections of exclusive tasks. Experiments on heterogeneous vehicle routing and open multi-depot pickup-and-delivery problems show that GeoPAR improves large-scale zero-shot generalization while substantially reducing rollout steps and maintaining efficient inference.
Read the full abstract
Artificial Intelligence
Consistency Without Alignment: Item-Sensitive Language Models Indistinguishable From Random
Item-sensitivity, defined as whether a model's choice depends on the specific input rather than on its own output prior, is widely reported as evidence of task competence. We show this evidence is necessary but not sufficient using a forced-choice signalling task abstracted from the board game Deception: Murder in Hong Kong. In this environment, the reference points against which a coordinate should be judged (a fit-maximising strategy, a posterior-maximising strategy, and uniform random selection) are all computable in closed form. Across seven language models, two model families, a post-training ablation, and three independent scoring rules, every one of 21 model-by-rule cells is reliably item-sensitive. Yet 8 of those 21 cells are not statistically distinguishable from a chooser that ignores the item and selects at random, and 5 score worse than random at describing the target. Item-sensitivity and distance from random correlate at only r = 0.30. We call this consistency without alignment and argue it generalises to any evaluation that relies on item-sensitivity, permutation consistency, or self-consistency without an independent reference for the measured quantity. We further find that a literal-similarity baseline with no pragmatics outperforms most tested language models, that adding a pragmatic layer over two baseline similarity sources moves choosers toward random rather than toward the Bayesian reference, and that a standard labelled multiple-choice format carries no measurable content signal here. All results represent the model side of a pre-registered instrument; a matched human condition is designed and piloted but not yet collected.
Read the full abstract
Artificial Intelligence
Residual Sparsification via Output Importance for Compressing Mixture-of-Experts LLMs
Mixture-of-experts (MoE) architectures scale large language models efficiently, but they demand massive GPU memory. To cope with such demand, models are commonly compressed to reduce their memory footprint. Residual sparsification is a representative compression technique that decomposes each projection matrix of an expert into a shared base matrix and per-expert residual matrix, and then compresses the residuals. Existing sparsification methods compress each residual matrix independently by minimizing its compression error, thereby minimizing the error of each projection matrix. However, our analysis shows that this objective is misaligned with preserving model accuracy after compression. In an expert, the final output is produced through computations coupled across multiple projections and hidden representations. Therefore, even small errors in individual matrices can propagate through hidden representations and projection interactions, leading to large expert output errors and accuracy degradation. To address this misalignment, we propose PARSER, a new residual sparsification method that shifts the compression objective from minimizing isolated matrix errors to preserving the expert output error. PARSER achieves this by introducing output importance, which measures the actual contribution to the expert output error. Our experiments show that, compared with existing methods, PARSER narrows the accuracy gap to the uncompressed model by 1.41$\times$ on Qwen and 1.44$\times$ on DeepSeek, while achieving the same peak memory reduction. Our code is available at https://github.com/OSSS-KU/PARSER.
Read the full abstract
Superconductivity
Expanding the trilayer Ruddlesden-Popper nickelate family: Synthesis and characterization of Sm$_4$Ni$_3$O$_{10-\delta}$ single crystals
The discovery of high-temperature superconductivity in Ruddlesden-Popper (RP) nickelates has attracted significant attention. Bulk superconductivity emerges under pressure in trilayer nickelates La$_4$Ni$_3$O$_{10-\delta}$ (T$_c$ $\approx$ 30 K) and Pr$_4$Ni$_3$O$_{10-\delta}$ (T$_c$ $\approx$ 40.5 K), where the reduced ionic radius of Pr$^{3+}$ may generate internal chemical pressure and enhance T$_c$. However, synthesizing trilayer RP phases with smaller rare-earth elements (Ln) is extremely challenging. So far, only the La, Pr, and Nd analogues have been synthesized with stable phases in the single rare-earth form. Here we report the first successful high-pressure and high-temperature (HPHT) synthesis of samarium-based compound Sm$_4$Ni$_3$O$_{10-\delta}$. Magnetization and transport measurements consistently confirm a density wave (DW) transition at ~180 K at ambient pressure. Through a careful fitting to the structural data of Sm$_4$Ni$_3$O$_{10-{\delta}}$, it is found that the bond angle of (Ni-O-Ni) associating with the interlayer apical oxygen is much smaller than 180$^{\circ}$, which was assumed to be the key factor for the occurrence of superconductivity. By applying pressures up to 80 GPa, despite partial suppression of insulating behavior and the DW order, but superconductivity is not observed in our present study. Density functional theory calculations suggest that the 3d$_{z^2}$ and 3d$_{x^2-y^2}$ are separated from other t$_{2g}$ orbitals and make a primary contribution to the Fermi surface. The newly synthesized trilayer nickelate Sm$_4$Ni$_3$O$_{10-\delta}$ offers a unique platform for probing the fundamental physics of RP nickelates.
Read the full abstract
General Relativity and Quantum Cosmology
Performance of quantum imaginarity in black hole spacetime
Quantum imaginarity is a fundamental quantum resource essential for quantum information processing. In this paper, we investigate the dynamical evolution behavior of quantum imaginarity for tripartite quantum mixed states in Schwarzschild spacetime. We find that Hawking radiation degrades the quantum resource of imaginarity, consistent with the behavior of entanglement and coherence under the Hawking effect. Interestingly, imaginarity exhibits a sharp change in the early stage of Hawking radiation, after which it varies slowly and asymptotically approaches a steady value. We further derive several trade-off relations that characterize the nonlocal redistribution of imaginarity across Rindler horizons. The findings of this study contribute to a deeper understanding of quantum resource behaviors in black hole spacetimes.
Read the full abstract
Computers and Society
A Mathematical Framework for Legacy, Governance, and Decision Integrity in Enterprise AI
Enterprise artificial intelligence is increasingly embedded in decisions that must remain lawful, explainable, adaptable, and accountable despite personnel turnover, model replacement, regulatory change, and shifting organizational incentives. Existing governance frameworks provide important principles but do not by themselves supply a compact mathematical language for evaluating whether an institution can preserve sound judgment over time. This paper develops a design-science framework for institutional legacy: the durable capacity of a decision system to continue producing beneficial, lawful, explainable, and adaptable outcomes after its original designers have stepped away. The framework contributes: (i) a normalized Legacy Score based on a penalized geometric mean of knowledge retention, governance, human oversight, adaptability, feedback learning, and jurisdictional fidelity; (ii) Decision Confidence and Decision Risk models separating evidentiary confidence from consequence; (iii) authority-aware retrieval and calibrated abstention; (iv) Decision Memory for governed organizational learning; (v) Regulatory Change Velocity mapping change exposure to review intervals; and (vi) a federated regulatory knowledge-graph architecture preserving provenance and legal hierarchy. The paper also proposes eight AI Decision Integrity Rules, an evaluation protocol, and a reproducible computational demonstration. The demonstration combines a deterministic stress test with 200 Monte Carlo replications of 10,000 synthetic decisions each, illustrating Legacy Score non-compensation and comparing consequence- and authority-aware routing with a matched-coverage confidence-only baseline. The contribution remains conceptual rather than field-validated; the simulation tests internal behavior, not production performance, and all parameters require context-specific calibration.
Read the full abstract
Differential Geometry
Shape Optimization and Bernoulli Free Boundary Problems for the Riemannian $p$-Laplacian
We study an exterior Bernoulli-type free boundary problem associated with the Riemannian $p$-Laplacian on a compact Riemannian manifold. Following the shape-optimization strategy of Ly and Seck, we formulate the corresponding volume-constrained shape optimization problem, derive its first shape derivative, obtain the overdetermined Bernoulli condition through a Lagrange multiplier, and establish a conditional monotonicity result under a Riemannian contraction hypothesis. The final successive-approximation argument will require additional geometric and boundary-flux stability assumptions.
Read the full abstract
Artificial Intelligence
VoiceLongMemEval: Do Assistants Remember How You Sounded?
With the growing scale of multi-agent architectures and large language models, deployed AI assistants are increasingly tasked with reasoning over long, continuous, multi-session conversation histories. Current benchmarks evaluate this dialogue history as information retrieval over long horizon, temporal reasoning, or knowledge updates, while crucially ignoring the fundamental dynamics of human-agent interaction, i.e. how they said it. To address this gap, we present VoiceLongMemEval (VLME) benchmark, where every answer depends on paralinguistic metadata (emotion labels, prosody descriptors, and voice events) attached to conversational turns, which is otherwise unrecoverable from the words alone. Every item passes a three-stage adversarial gate, ensuring that a strong language model fails when given only the transcript. Evaluating leading frontier and open-weight models reveals a pervasive affect gap; providing text-track paralinguistic metadata yields a 0.09 to 0.38 accuracy boost (0.61 to 0.69 when prompted with evidence hints), while standard ASR pipelines systematically discard this signal. Additionally, audio-native models successfully extract these cues directly from speech (0.354 to 0.412 vs. 0.325 blind). Code and dataset will be made available upon acceptance.
Read the full abstract
Game Theory
Graph Coloring with Color Preferences
We study graph coloring with color preferences, in which each vertex ranks the available colors. In addition to assigning different colors to adjacent vertices, we require the coloring to be stable: no group of vertices can cyclically exchange their assigned colors so that each strictly prefers its new color to its original one. We define the stable chromatic number $\chi_\mathrm{stable}(G)$ of a graph $G$ as the minimum integer $k$ such that every preference profile admits a stable $k$-coloring of $G$. We establish several upper and lower bounds. In particular, for any acyclic orientation of the edges of $G$, the largest number of vertices reachable from a vertex by directed paths, including the vertex itself, is an upper bound on $\chi_\mathrm{stable}(G)$. This shows that $\chi_\mathrm{stable}(G)$ is well-defined. We also show that $O(t \log (1+n/t))$ colors suffice for an $n$-vertex graph $G$ of treewidth $t$, and complement this with a lower bound in terms of the Grundy number. Turning to the problem of finding a minimum stable coloring for a given profile, we show that stable $2$-colorability is polynomial-time solvable, whereas stable $k$-colorability is NP-complete for every fixed $k\ge 3$. Using the treewidth bound, we give a fixed-parameter tractable algorithm parameterized by treewidth.
Read the full abstract
Software Engineering
WiseSpec: Requirements-Driven Agents for Code Generation
Code generation aims to automatically generate source code from task requirements and has attracted significant attention with the rapid advancement of large language models (LLMs). Despite remarkable progress, LLMs often struggle to generate correct code for complex software engineering tasks because task descriptions are frequently incomplete, ambiguous, or lack critical contextual information. Existing approaches primarily improve the capabilities of coding agents through more sophisticated tools, skills, and workflows, while largely overlooking the quality of the task requirements themselves. To address this limitation, we draw inspiration from software requirements engineering and propose WiseSpec, a novel requirements-driven agent framework for repository-level code generation. WiseSpec automatically constructs structured and information-rich requirements, assesses their quality through execution-based evaluation, and iteratively refines them to better guide code generation. Experimental results show that WiseSpec consistently outperforms all baselines, achieving an average improvement of 13.17% in %Resolved.
Read the full abstract
Algebraic Geometry
Hodge–Tate splitting and Akizuki–Nakano vanishing in positive characteristic
We introduce the notion of Hodge–Tate splitting for schemes in positive characteristic. For a smooth variety $X$ over a perfect field $k$ of characteristic $p > 0$, we say that $X$ is Hodge–Tate split if the natural morphism $\mathcal{O}_X \to F_*\Omega^\bullet_{X/k}$ induced by the absolute Frobenius admits a splitting in $\mathcal{D}_{\mathrm{qcoh}}(X)$. We prove that this condition is equivalent to the existence of a decomposition $F_*\Omega^\bullet_{X/k} \simeq \bigoplus_{i=0}^{\dim X}\Omega^i_{X/k}[-i]$ of its de Rham complex. Consequently, for smooth projective varieties, Hodge–Tate splitting implies Akizuki–Nakano vanishing and the $E_1$-degeneration of the Hodge-to-de Rham spectral sequence. Furthermore, we establish criteria and permanence properties for Hodge–Tate splitting and use them to construct many new examples of varieties whose de Rham complexes decompose. These include blow-ups of quasi-$F$-split varieties along strata of simple normal crossings divisors, complete intersections in toric varieties, and linearly reductive quotients of Hodge–Tate split varieties. Among these examples, we obtain smooth projective varieties whose Hodge-to-de Rham spectral sequences degenerate at $E_1$, whereas their Hochschild–Kostant–Rosenberg spectral sequences do not degenerate. As an application in mixed characteristic, we prove an Akizuki–Nakano-type vanishing theorem for smooth projective globally $+$-regular varieties over the Witt ring of a perfect field.
Read the full abstract
Machine Learning
EEG-VID: Task-Guided Latent Predictive Pretraining for EEG Decoding and Assistive Target Selection
We propose EEG-VID, a task-guided latent predictive pretraining framework for EEG decoding under session and subject shifts. EEG-VID predicts future latent EEG states from recent history using an exponential-moving-average target encoder and weak task guidance, followed by supervised fine-tuning. Across VIG-48 and BCI Competition IV-2a/IV-2b, Stage 1 improves mean accuracy in 41 of 42 matched backbone-dataset-protocol comparisons, including all 12 leave-one-subject-out settings, with a maximum gain of 16.22 percentage points. On the 48-region cross-day VIG-48 task, EEG-VID achieves 6.52% Top-1 and 30.50% Top-5 accuracy. In a separate six-participant offline robot-scene study, candidate-constrained target selection reaches 40.24% versus a 25% chance level after subject-specific calibration. These results support task-guided latent prediction as a transferable pretraining strategy for EEG decoding and scene-constrained assistive target selection.
Read the full abstract
Social and Information Networks
Aligned but Flattened: Analyzing the Trade-off between Cultural Alignment and Diversity in LLMs
Cultural fine-tuning has become the de facto paradigm for building culture-aware large language models (LLMs), yet existing optimization exclusively for alignment scores provides an incomplete portrait of cultural fidelity by systematically obscuring inherent cultural diversity. This unidimensional evaluation lens prompts a fundamental question: do models genuinely perceive distinct cultural nuances, or do they merely memorize dominant cultural values? To address this, we propose a synergistic evaluation framework that jointly formalizes cultural alignment and diversity. Through extensive benchmarking of six mainstream LLMs on the World Values Survey, this framework uncovers a systematic and critical trade-off: the pursuit of cultural alignment consistently incurs an acute expense of diversity, leading to severe "cultural flattening." Investigating this behavioral shift, we demonstrate that these superficial alignment gains stem from models artificially anchoring to dominant majorities, converging onto a monolithic response pattern that wipes out the heterogeneous distributions inherent to human groups. Crucially, our mechanistic analysis suggests that this diversity collapse is not merely a behavioral anomaly but more likely a structural consequence of the low-rank bias inherent in neural network optimization. Therefore, our findings expose the limitations of current post-training paradigms and call for a shift toward alignment objectives that preserve cross-cultural pluralism.
Read the full abstract
Biological Physics
Mudskippers use tail thrusting to help crutching to move on mud of various wetness
At the water-land interface, amphibious fishes encounter wet flowable substrates made of granular solid-water mixtures, which can stay solid or flow like a fluid. As these substrates become wetter or drier, their yield strength (at which solid-fluid transition occurs) and cohesion (how sticky they are) both change, challenging locomotion. Despite substantial understanding of tetrapod locomotion on flowable substrates (mostly dry sand), we know little about how amphibious fishes cope with wet flowable substrates of various wetness. Here, we studied mudskippers on clay mud of controlled, variable wetness over the range where solid-fluid transition occurs. As mud became wetter, its strength decreased by 100-fold, leading the animal to sink deeper, with larger areas of body and fins contacting mud. By contrast, mud stuck most easily at intermediate wetness. The increased sinkage and contact and stickiness change caused more mud to stick to and pull against the animal on wetter mud. We also tested dry mud, which stuck to animal fins as its mucus dried. Despite these challenges, the mudskipper predominately used a conserved crutching gait on all except the wettest mud tested, with a modest performance reduction. When normal crutching became less effective, the animal assisted it with tail thrusting, by bending and straightening it to push downward and backward to generate additional thrust and lift, or even thrusting the tail to jump. These observations suggest that mudskipper's crutching motor program is well adapted to its native muddy substrates but inflexible, with most novelty in tail use.
Read the full abstract
Biological Physics
Importance and methods to control, vary, and characterize mud strength for studying locomotion
Animals and robots encounter mud at the water-land interface. Like sand, mud can stay solid or flow like a fluid. Unlike sand, the yield strength of mud at which solid-fluid transitions occur depends on not only the amount of solid relative to fluid (water in mud, air in dry sand), but also how much coarse grains and fine clay are within the solid. Despite understanding of locomotion on/within dry sand dominated by coarse grains with repulsive normal forces and friction, little is known for mud dominated by fine clay with strong cohesion. Here, we developed methods to prepare uniform mud of controlled, variable yield strength and characterize and track its drift from water evaporation. Compared to other flowable substrates, mud strength measured by upward force during penetration is weaker and can vary more, and mud sticks more during extraction to pull downward, making it more challenging for locomotion.
Read the full abstract
Sound
BiMTokenizer: Preserving Semantic-Acoustic Balance in Low-Bitrate Speech Tokenization via Bidirectional State-Space Modeling
Speech codecs serve as bridges between continuous speech signals and large language models, yet face an inherent conflict between acoustic fidelity and semantic preservation. To mitigate this conflict, recent works increasingly adopt dual-tower architectures to decouple semantic and acoustic modeling with separate encoders. However, these dual-tower designs incur substantial architectural overhead. To avoid such complexity, we revisit the single-tower paradigm and propose BiMTokenizer, a low-bitrate speech codec (around 1.1 kbps) combining a bidirectional state-space backbone with Residual Spherical Leech Quantization (RSLQ). The bidirectional backbone strengthens temporal modeling, while RSLQ offers a fixed, well-separated lattice bottleneck for robust semantic and acoustic tokenization without learned-codebook collapse. Experiments show that BiMTokenizer achieves superior acoustic reconstruction and the lowest WER among low-bitrate codec baselines across both clean and noisy environments, while using less than half the parameters of recent dual-tower baselines. Furthermore, its robust semantic representations yield strong performance on downstream speech understanding tasks, confirming that a well-designed single-tower codec can preserve the semantic-acoustic balance at low bitrates. The code and model weights are available at https://github.com/ZhangXinWhut/BiMTokenizer.