Peer Revenue Returns subject to peer review

Papers

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

All of arXiv

Monday 9 November

Cryptography and Security

Splitting for Stability: a Modified Reward Scheme for Transaction-Fee-Only Regimes

As Bitcoin's block subsidy declines over time, miner revenue will come to depend increasingly on transaction fees. In fee-dominated regimes, winner-take-all reward structures can create incentives for undercutting and fork-based mining strategies, leading to instability and reduced transaction throughput. To prevent this instability, we propose a reward-splitting mechanism in which the winning block producer receives an $\alpha$-portion of the block reward, while the rest is distributed among other miners; this incentivizes mining on previous blocks with higher rewards, while preserving the incentive to produce blocks with high rewards. We evaluate the proposed reward structure through analysis and simulation. Our analysis derives a local profitability condition for one-step Function-Fork deviations, showing how the winner-side reward, loser-side subsidy, and honest hash power jointly determine whether undercutting is profitable. Simulations show parameter regions in which honest behavior dominates fork-based deviations. In the cases where forking becomes dominant, the learned strategy is the least aggressive variant studied. The analysis and simulations suggest that suitable reward-splitting parameters can reduce the honest participation needed to suppress aggressive forking strategies relative to the winner-take-all structure studied in prior work. Finally, we show that the modified reward scheme does not appear to worsen selfish mining incentives. Together, these results suggest that reward redistribution can mitigate fee-driven fork incentives without introducing stronger selfish mining incentives, providing a possible direction for incentive design in fee-era proof-of-work systems.
Read the full abstract

Robotics

Curie's Jarvis: Autonomous Robotic Chemistry Experiments from Printed Protocols

A person running a bench experiment works from a printed instruction sheet, tells identical containers apart by the labels written on them, and looks at the bench again before each step. Lab automation usually works the other way around: the procedure is written as a script, and containers stand at fixed positions or carry fiducial markers. We present Curie's Jarvis, a system that lets a robot arm carry out a chemistry experiment on a bench set up for people. It reads the steps from a photo of the printed sheet, finds containers on an ordinary bench with no fixed positions or markers, tells identical-looking containers apart by reading their labels in four camera views, locates its target again before every step, and sets how each step is carried out, including its speed, from what is being handled. On a Franka Panda arm, Curie's Jarvis completed 5 of 5 runs of one protocol from a children's chemistry kit and 4 of 5 runs of another, with no human intervention, and in 160 locates on printed and handwritten labels it identified every container correctly. This brings robots a step closer to doing bench chemistry the way people do.
Read the full abstract

All authors Rumi Loghmani, Bozhang Song, Victor Alvarado, Aliyah Redding, Amir Barati Farimani

Software Engineering

Using Proof Dependencies to Manage Verification Projects

Verification technology now scales to sizable projects, but techniques for managing such projects fall short. There are no effective ways to assess the progress made so far and to determine which verification tasks to perform next. A key issue is that the successful run of a verification tool provides little indication about what has actually been achieved. The guarantees provided depend on the specifications of the verified code as well as any unverified assumptions made in the proofs, e.g., about the behavior of unverified libraries. Manual inspection of specifications and assumptions to determine progress is error-prone and prohibitively expensive. In this paper, we present automated techniques to assess and guide verification projects. All of these techniques employ an automatic proof dependency analysis that computes, for each proof obligation in a project, the specifications, statements, and assumptions that are needed to prove it. We show that this analysis enables effective proof audits by determining all assumptions a proof depends on. We also use it to determine which statements in a program are constrained by the provided specifications, which indicates how strong a specification is. Based on these two insights, we propose a novel progress metric for verification projects and show how to use it to guide verification projects to maximize progress. We implement our approach in the symbolic execution backend of Viper and lift it to the Go verifier Gobra. Our evaluation on real-world benchmarks, including a verified Internet router, shows that the dependency analysis is sound and sufficiently precise, that our progress metric accurately reflects the state of a verification project over time, and that our guidance outperforms standard bottom-up and top-down verification strategies.
Read the full abstract

All authors Lea Salome Brugger, Andrea Keusch, João Pereira, Peter Müller

Machine Learning

Domain-Routed Delta Optimization for Multi-Teacher On-Policy Distillation

Multi-teacher on-policy distillation can yield uneven domain capabilities. We propose Domain-Routed Delta Optimization (DRD-MOPD), composing a frozen base with expert-initialized trainable deltas. Domains update their deltas with independent optimizer states and clipping. After joint training, Post-separate exports the base plus one trained delta per domain, selected by known task domain without retraining. Using AdamW, the unweighted full student reaches 32.83% Macro across six benchmarks after 20 updates, versus 32.21% for a matched shared control. Post-separate raises Macro with Open-MOPD weighting.
Read the full abstract

K-Theory and Homology

Hermitian K-theory for stable $\infty$-categories IV: Karoubi-Grothendieck-Witt theory

In analogy with the passage from connective to non-connective algebraic K-theory, Schlichting put forward the Karoubi-Grothendieck-Witt spectrum as an approximation to Grothendieck-Witt theory by a functor invariant under idempotent completion. In this paper, we extend his construction to the setting of Poincaré $\infty$-categories and characterise the resulting functor by the expected universal property. Our main result places these Karoubi-Grothendieck-Witt-spectra in an exact sequence involving the ultimately decorated L-spectrum and homotopy orbits of non-connective K-theory. This confirms a conjecture of Williams which had remained open even for rings in which $2$ is invertible, and allows us to deduce universal characterisations for ultimately decorated L-theory. We obtain these results by unifying the Bass-Heller-Swan and Ranicki-Shaneson splittings into a general decomposition theorem that yields, in particular, formulas for Grothendieck-Witt groups of Laurent polynomial rings. Finally, we construct multiplicative structures on both versions of our Grothendieck-Witt and L-functors and refine their universal properties to multiplicative ones. Along the way, we show that the original versions are even idempotent with respect to Day convolution.
Read the full abstract

All authors Baptiste Calmès, Emanuele Dotto, Yonatan Harpaz, Fabian Hebestreit, Markus Land, Kristian Moi, Denis Nardin, Thomas Nikolaus, Wolfgang Steimle

High Energy Astrophysical Phenomena

Detectability of Seyfert galaxies in point-source and stacking searches with a global network of neutrino telescopes

Recent evidence for high-energy neutrino emission from NGC 1068 and other nearby Seyfert galaxies suggests that active galactic nuclei (AGNs) may constitute a class of astrophysical neutrino sources. Using a physically motivated neutrino emission model based on stochastic particle acceleration in magnetized AGN coronae, we investigate the detection prospects of X-ray-bright Seyfert galaxies with a detector network comprising IceCube, Baikal-GVD, KM3NeT/ARCA, and P-ONE. We find that the Circinus Galaxy is the most promising point-source target in the Southern sky and will be detected at the 5$\sigma$ level by 2038 at the latest, provided that the Seyfert population accounts for the diffuse astrophysical neutrino flux up to 100 TeV and Circinus is at least as efficient a neutrino emitter as the average source in the population. Individual detections of other nearby Seyferts are expected to remain challenging, making stacking searches informed by X-ray observations the most promising approach for discovering them within the next decade. Our results demonstrate that joint analyses across multiple neutrino telescopes, together with future MeV gamma-ray observations, will be essential for establishing Seyfert galaxies as astrophysical neutrino sources and provide a means for testing the coronal origin of their neutrino emission.
Read the full abstract

Artificial Intelligence

ExpertCrowd-SWE: Measuring Frontier Coding Agents on Expert-Authored Engineering Tasks in Unfamiliar Codebases

ExpertCrowd-SWE (EC-SWE) is a benchmark of 100 expert-authored, multi-file software engineering tasks for evaluating autonomous coding agents in large, unfamiliar codebases. Each task is a change request written by a practising engineer against a pinned commit of a real repository, paired with a sealed reference solution and a task-specific grader whose weighted, mostly executable criteria award partial credit; must-pass blockers and pass/fail gates encode what a maintainer would refuse to merge. The tasks span 87 open-source projects and 11 languages. The median ticket is 148 words, while the median reference solution changes 471 lines. Every task image is calibrated before release so that the reference solution scores 1.0 and both the untouched repository and an inert change score 0; nearly half of the tasks also ship alternative, lazy and adversarial control patches that the grader must classify correctly. The final 100 tasks were selected for difficulty against current frontier models. We evaluate seven frontier models, five runs each (3,495 scored runs), in a single platform harness that gives each model its provider-native tool interface. Claude Opus 5.5 (67.3%) and GPT-6 Astra (63.7%) lead without a statistically significant gap, followed by Claude Fable 5.1 (46.5%), Gemini 3.8 Flash (40.0%), and a statistically indistinguishable group of Qwen 3.8 Max, DeepSeek V4.1 Flash and Grok 4.6 (32-34%). No task is fully solved by every model, and on 80 tasks no model earns full credit on all five runs. Scores span 35 points. Depending on the model, agents take a median of 33 to 166 steps per run, with little relation to score. We release 24 tasks with their environments, graders and results, and hold out 76 to limit contamination.
Read the full abstract

All authors Josip Vrdoljak, Akash Singh, Bas Steunebrink, Alex Yarotsky, John Lorenzo, Achille Nicoletti, Sam Pigott

Computer Vision

Rethinking Evaluation for Video Capsule Endoscopy: Validation Protocol, Source-Video Diversity, and Cross-Dataset Shift

Automated video capsule endoscopy (VCE) classification remains challenging due to severe class imbalance, limited source-video diversity, and variations in image appearance across datasets. Although deep learning models have demonstrated promising performance on established benchmarks, their reliability under different validation protocols and external dataset shift remains insufficiently understood. In this work, we introduce DMSNet, a deformable adaptive multi-scale edge-guided attention network integrating adaptive multi-scale feature fusion, deformable lesion refinement, edge-guided attention, and sparse feature selection. We benchmark DMSNet against 18 representative convolutional, transformer-based, and pretrained vision architectures on Kvasir-Capsule under random frame-level and source-video-disjoint validation protocols. We further identify source-video overlap between the original training and test splits and evaluate model performance on both the complete test set and a source-video-disjoint subset. Under random frame-level validation, DMSNet achieves an MCC of 0.442 $\pm$ 0.020 on complete test set and 0.456 $\pm$ 0.037 on the source-video-disjoint subset. On the external CV2024 dataset, ConvNeXt-Tiny achieves the highest MCC of 0.207, while performance degrades substantially across architectures relative to in-domain evaluation. We further investigate source-video diversity using a per-class cap of 200 training frames. With the total training-set size held fixed, increasing the maximum source-video coverage from four to eight improves MCC for 17 of 19 evaluated architectures. Overall, these findings demonstrate that model rankings and classification performance depend substantially on evaluation protocol, training-source diversity, and cross-dataset shift, emphasizing the need for source-aware validation and rigorous external evaluation of VCE classification systems.
Read the full abstract

Combinatorics

Stabilization of the zero forcing number of generalized Petersen graphs

We show that the zero forcing number of the generalized Petersen graph $P(n,k)$ is $2k+2$ whenever $n\ge 2k^2+k+1$. We prove this by showing that the zero forcing number of the infinite generalized Petersen graph $P(\infty,k)$ is $2k+2$ and bounding the width of the closure of every set of at most $2k+1$ vertices. This bound gives a threshold on $n$ beyond which the projection of such a closure onto $P(n,k)$ does not wrap around. Combined with an exhaustive computer search for smaller $n$, we determine the zero forcing number of every $P(n,k)$ for $k\le 5$.
Read the full abstract

Machine Learning

GigaEdge Spench: An SDK for Practical Streaming Speech Enhancement

We present GigaEdge Spench, an open-source SDK for streaming speech enhancement on CPU servers, mobile devices, and in browsers. Applications can build enhancement channels from neural denoising, normalization and dynamics blocks, sample-rate conversion, and acoustic echo cancellation where supported. These stages can be combined, while normalization can also run without a neural model. The release includes an 8 kHz INT8 denoiser and discriminative and hybrid 48 kHz FP32 models. A shared C++ core supports ONNX Runtime and TensorFlow Lite through native interfaces, Android and iOS SDKs, and WebAssembly integration. Independent stream contexts share model resources, and consecutive frames from one stream can be processed in a single inference call. We evaluate the released models on EARS-WHAM v2 and compare their quality and processing costs with RNNoise and DPDFNet2. All tested native and Web configurations process files faster than real time, although larger chunks require more time to collect input. We also measure how many streams a shared model can process on a CPU. Our contribution is a configurable deployment system and an experimental study of how model choice, chunk size, and execution environment affect its use. All reported SDK timing uses ONNX Runtime. Index Terms: speech enhancement, streaming inference, cross-platform deployment, temporal frame grouping, computational efficiency
Read the full abstract

All authors Artem Sokolov, Artem Khaidarov, Yana Lysova, Alexander Gladkikh, Vitaliy Yulis

Optics

Molecular Gas Optics Using Vibrational Population Density Refraction

Gas optics holds promise for development of advanced optical elements with 2-3 orders of magnitude higher damage threshold than any solid-state optics. Here, we demonstrate a new concept for refracting the laser beam using vibrationally excited molecules as gas optics instead of traditional control of thermodynamic gas temperature. The transient, vibrational population density grating is formed in a CO2 gas by interfering two mid-IR lasers to imprint spatial modulation in the linear refractive index (Delta n0 >=2x10-5) of the neutral gas via enhancement of the molecular polarizability. It is increased by pumping the asymmetric stretching mode of CO2 at optimal wavelengths to resonantly produce a highly nonequilibrium state that increases the effective vibrational mode temperature to greater than 2000 K in a room temperature gas. The measured beam steering efficiency of such a volume gas grating exceeds 90% that makes this approach a candidate for producing other large-aperture, high-throughput molecular optics elements, with expected damage thresholds 0.5-1 kJ/cm2, much-needed for future ultrahigh-power lasers.
Read the full abstract

Software Engineering

FlowPlan: Reducing the Token Cost of Coding Agents with Abstraction-Guided Planning and Verification

LLM-based coding agents such as Claude Code can resolve real-world repository issues, but their iterative interaction loop can be expensive. Existing approaches reduce this cost primarily by limiting, compressing, or selectively avoiding information and actions within the repair process. We investigate a complementary question: Can an agent reason about a repair at an abstract program level, rather than repeatedly operating over raw source code, to reduce both the exploration needed to identify a repair and the repeated checking needed to verify it afterward? We present FlowPlan, an abstraction-guided repair approach that structures localization, planning, and verification around per-function control-flow or data-flow graphs. The agent first localizes the candidate fault location on the program abstraction and expresses the intended repair as explicit graph deltas before modifying the concrete source. After the edit, FlowPlan reconstructs the same abstraction and compares the observed structural change with the planned one. Then, a verification checks the complete repair before termination. Across all six benchmark-model settings, FlowPlan does not significantly change resolution rate relative to the vanilla agent while reducing token consumption in every setting, significantly in five of six settings. With Claude Haiku 4.5, it uses only 0.46-0.47x the baseline tokens on SWE-bench Pro, 0.32-0.35x on SWE-bench Verified, and 0.68-0.74x on SecBench; with Sonnet 5, it uses 0.72-0.76x, 0.73-0.74x, and 0.92-0.94x, respectively. Overall, FlowPlan achieves the largest token reduction among the compared methods in three of the six settings and ties the strongest compared method in a fourth. Trajectory analysis shows that FlowPlan saves tokens by replacing much of the costly post-edit re-reading and testing with explicit verification and reducing per-call context.
Read the full abstract

All authors Melika Sepidband, Jiho Shin, Song Wang, Hadi Hemmati

Statistics Theory

Local Fréchet regression with polyspherical predictors: Uniform convergence rates and factor selection

We study local Fréchet regression for responses taking values in a metric space and predictors on a product of spheres of possibly different dimensions. We propose local constant and local linear estimators with factor-specific bandwidths and establish uniform convergence rates. We select sphere factors by comparing cross-fitted prediction risks for the full and leave-one-factor-out models. The procedure consistently recovers the active factor set, and the selected local constant and local linear estimators attain the same uniform convergence rates as if the active factor set were known. For univariate Wasserstein responses supported on a common compact interval, the uniform rate holds without an entropy condition on the response space. Simulations with scalar and spherical responses and analyses of vessel speeds and drifter speed distributions illustrate the methods.
Read the full abstract

Representation Theory

A decomposition theorem for minimal W-algebras of type A

To any simple Lie algebra, one can associate a minimal affine W-algebra. It is a vertex algebra which is obtained from the universal affine vertex algebra by a chiral Hamiltonian reduction procedure. This reduction is determined by data coming from the minimal nilpotent orbit. It is a chiral quantisation of the Slodowy slice associated to this minimal orbit. In this paper, when the simple Lie algebra is of type A, we prove that the minimal W-algebra is isomorphic to a commutant subalgebra in a localisation of the universal affine vertex algebra. Then, we establish a decomposition theorem à la Losev: the localised universal affine vertex algebra decomposes as the tensor product of the minimal affine W-algebra and another vertex subalgebra. As a direct corollary, we retrieve the celebrated result of inverse Hamiltonian reduction for the minimal W-algebra. We also prove an analogue of this result for the corresponding finite W-algebra. In both cases, our approach relies on the geometry of the associated Slodowy slice.
Read the full abstract

Signal Processing

STAR: Semantic-Criticality Tiered Allocation of Radio Resources via Zero-Shot Learning

In remote inference, edge devices transmit high-resolution images over wireless links to a receiver, where an artificial intelligence (AI) model answers task-specific questions about these images to enable reliable decisions. These answers are often determined by a small set of task-causal pixels, such as those of text, whose corruption changes the answer, so these pixels must be delivered with high fidelity. Conventional transmission protects every pixel equally, so it either consumes many resource elements (REs) or corrupts these regions at low signal-to-noise ratio (SNR). To overcome these limitations, in this paper, a zero-shot, domain-agnostic framework, termed semantic-criticality tiered allocation of resources (STAR), is proposed. In particular, a per-pixel semantic-criticality indicator (SCI) is derived from the attention of a frozen vision-language model, so new tasks and domains are handled by changing only the prompt, without retraining. The resulting task causality-aware resource allocation problem over protection tiers is solved exactly by a threshold rule on the SCI, realized by a coded tier for the task-causal pixels and an uncoded, downsampled base layer for the full image. Simulations over a 5G New Radio link demonstrate that STAR attains a semantic accuracy of at least 0.98 for $E_s/N_0 \geq -4$ dB and, at comparable semantic reliability, reduces the REs of robust coded transmission by 59.5% and 91.5% on factory- and vehicle-monitoring datasets, respectively, the latter without any change to the models. Although a deep joint source-channel coding baseline uses fewer REs at comparable semantic accuracy, it must be retrained with labeled data for every new task or domain.
Read the full abstract

Computers and Society

Understanding Computing Instructors' Perceptions of Student Sense of Belonging

Students' sense of belonging has been described as a feeling of community acceptance, respect, inclusion, and membership. In computing, student belonging is shaped by psychological, social, academic, and environmental factors, including their relationships with instructors. However, despite prior work emphasizing instructors' influence on student belonging, limited research has examined how computing instructors perceive, understand, and attempt to support belonging. Through nine semi-structured qualitative interviews with introductory computing instructors at large public universities in the United States, we investigated how instructors define, measure, and conceptualize belonging; what interventions they use to support student belonging; and what barriers limit the efficacy of these interventions. Our findings show that instructors defined belonging through value and identity representation, self-efficacy, and community; relied on surveys, observations, and student performance to measure belonging; saw belonging to be influenced by prior experience, major, student and institutional culture, and representation; implemented student-oriented interventions and designed courses around belonging; and identified multi-level challenges to supporting belonging. These results highlight opportunities to design belonging interventions that attend to both prior work and instructor conceptualizations, while also considering tensions in utilizing teaching assistants, balancing students' expectations, and adapting to broader societal trends affecting belonging.
Read the full abstract

All authors Ritesh Kanchi, Teresa Luo, Sam Grunebaum, Alex Oesterling

Artificial Intelligence

NCP: Verifier-Guided Optimization of Weight-Space Control Policies

We present the Neural Control Plane (NCP), a closed-loop system that observes language-model execution, synthesizes candidate weight-space intervention rules, evaluates them, and maintains a versioned table of admitted policies. NCP separates a pretrained model's inference data plane from the control policy that selects temporary parameter interventions. W-ISA supplies transactional execution and parameter restoration; NCP supplies telemetry, candidate synthesis, admission criteria, priorities, and rule lifetime management. The optimization state consists of active rules and an archive of evaluated candidates, while base-model parameters are restored after each request. In a local SmolLM-135M-Instruct experiment, NCP generated three candidates from development telemetry and evaluated them on separate task examples and a 4,095-target WikiText-2 slice. One projection reduced validation token repetition from 0.4133 to 0.2191, with a perplexity change of -0.0574%. All candidates produced zero newly correct exact answers, so the configured admission gate archived all three and retained the original active policy. Fresh-process repetition and saved-table resumption reproduced the measured outcomes. NCP provides an executable framework for studying recursive self-improvement at the control-policy level through explicit synthesis, verification, and state transitions.
Read the full abstract

Mesoscale and Nanoscale Physics

Neutron Scattering studies of the density dependence of One-dimensional Quantum Fluids

In previous work, we demonstrated that $^4$He confined within a templated porous material preplated with a noble gas to enhance dimensional reduction exhibits one-dimensional behavior. Neutron scattering revealed that the excitations in confined $^4$He are qualitatively distinct from those in two- and three-dimensional superfluid helium, aligning with quantum Monte Carlo calculations. These simulations suggest that three solid layers form, with a one-dimensional liquid located at the center of the pore. Here, we present neutron scattering results as the filling fraction of the one-dimensional core is varied. We observe that the one-dimensional liquid always forms with a fixed density, even as the pore filling increases. We posit that this is due to the growth of the one dimensional extent of the liquid as the filling is increased, but its density remains unchanged.
Read the full abstract

All authors Sutirtha Paul, Rahul Soni, Garfield Warren, Adrian Del Maestro, Paul E. Sokol

Nuclear Theory

The Alpha Clustering Model of the EMC Effect

For more than forty years, the EMC effect, the observation that nuclear structure functions in deep-inelastic scattering appear to be modified relative to the free-nucleon expectation, has resisted a consensus microscopic explanation. In this letter, we argue that the strength and $A$ dependence of the EMC effect may be governed by the phenomena of $\alpha$ clustering in nuclei and responsible for a majority of the partonic modification associated with the EMC Effect. We motivate the alpha clustering model of the EMC effect (ACME), and with a data driven approach, we demonstrate how such a model can reproduce the existing deep inelastic scattering data on nuclei. We make predictions for other nuclei such as $^{16}$O, $^{20}$Ne, and $^{48}$Ca and explore its consequences for a spin-dependent EMC effect and global analysis of nuclear parton distributions.
Read the full abstract

Artificial Intelligence

Decisions by Discrimination: Differential Evidence for Medical Reasoning

Large language models (LLMs) are increasingly studied for clinical decision support, and medical question answering provides a testbed for evaluating this capability. In particular, medical multiple-choice questions (MCQs) reflect a core challenge in clinical decision-making: plausible options often share overlapping evidence, while only a few decisive factors distinguish the correct option from incorrect ones. However, existing retrieval-based and reasoning-oriented methods mainly focus on supplying broadly relevant knowledge or eliciting reasoning paths, rather than identifying the compact differential evidence that separates confusing alternatives. In this work, we present a retrieval-grounded differential evidence framework for medical MCQs. Our method extracts option-specific pieces of differential evidence and organizes them into concise evidence prompts for answer selection. This design better aligns external knowledge with differential reasoning while reducing irrelevant context. Experiments show consistent accuracy gains across multiple datasets and target LLMs. Ablation studies further show that our framework outperforms both directly using retrieved evidence chunks and closed-source LLM-generated general hints, and that its improvements are not solely explained by answer-proximal guidance, but are consistent with differential evidence serving as a compact aid for reasoning over competing alternatives. Our findings also offer insights for medical reasoning and other decision-making settings with multiple possible choices.
Read the full abstract

All authors Yunpeng Xiao, Yuzhang Xie, Junhao Huang, Carl Yang, Xiao Hu, Kai Shu

Artificial Intelligence

W-ISA: Transactional Execution of Weight-Space Intervention Programs

We present W-ISA, a software instruction set and transactional runtime for executing weight-space intervention programs on pretrained language models. Shape-checked instructions express parameter updates, projections, ablations, and rank-one transformations. The runtime applies these programs within request-scoped execution contexts and restores saved parameters after completion or interruption. Activation-based match-action rules connect measured model states to intervention selection. We evaluate restoration and replay on SmolLM-135M-Instruct and Qwen2.5-1.5B-Instruct, and compare materialized rank-one updates with equivalent PyTorch hooks on the latter. In 45 timed requests, both intervention implementations produce identical greedy sequences at 32, 128, and 512 tokens; their maximum prompt-logit difference is 0.000184. On the complete WikiText-2 raw test split, the combined SmolLM attention-head ablation and MLP projection changes perplexity from 22.2353 to 22.3188, a relative increase of 0.3754%. W-ISA separates intervention specification, execution, and measurement, making parameter-level experiments reproducible and composable.
Read the full abstract

Numerical Analysis

Block-Sparse Gauss-Newton Optimization of Finite Basis Physics-Informed Neural Networks: Sparse Automatic Differentiation Approach

In this work, we develop an efficient Gauss-Newton optimization approach for finite basis physics-informed neural networks (FBPINNs). The key ingredient is the overlapping domain decomposition architecture of FBPINNs, which localizes the influence of each subnetwork and induces block sparsity in both the residual Jacobian and the corresponding Gauss-Newton matrix. We characterize this structure and exploit it through sparse automatic differentiation, using graph coloring to partition structurally orthogonal Jacobian columns or rows. This reduces differentiation costs and memory requirements while preserving all structurally nonzero derivative information. The resulting compressed Jacobian representations allow us to assemble the Gauss-Newton system directly in block-sparse form, enabling the use of sparse direct solvers and preconditioned iterative methods. In this way, the otherwise expensive second-order optimization problem is reduced to scalable sparse linear algebra. We demonstrate the approach on multiscale PDEs, where accurate approximation requires rich neural network parametrizations and large numbers of collocation points. Numerical experiments show that our block-sparse Gauss-Newton method attains high accuracy and substantially accelerates optimization relative to first-order baselines while remaining computationally tractable for large problem sizes.
Read the full abstract

General Economics

Dynamic investment and coordination in teams: Theory and experimental evidence

Teamwork requires more than the general skills that team members bring. In successful teams, members invest in team-specific skill or capital that accumulates within the team and raises the return to coordinated production. We provide the first systematic analysis of dynamic investment in such team skill. Adopting a dynamic game approach, we develop a novel theoretical framework where investment in team skill creates persistent benefits but is risky because the benefits depend on successful team coordination. We then take this framework to the laboratory to study empirically the factors that influence dynamic investment in team skill. We find underinvestment compared to the efficient benchmark. However, investment responds strongly to incentives, in line with specific theory-predicted patterns. Pre-registered traits (theory of mind and propensity to coordinate) predict investment and team-level efficiency. We conclude that careful design of team incentives and selection of team members can facilitate the development of team skill.
Read the full abstract

Robotics

Incorporating World Models in Decision Making for Next Best View Selection in Active Sensing

Efficient robotic 3D reconstruction requires selecting views that contribute new surface measurements with few physical captures. Pretrained world models can guide this choice, but a predicted surface may not produce valid depth measurements under the operating conditions of the deployed sensor. We present a deployment adaptation methodology for next-best-view selection that learns a sensor-return adapter from limited local calibration data while keeping the pretrained predictor fixed. The adapter estimates measurement dropout over each predicted surface, allowing the selection method to favor views likely to fill gaps in the measured reconstruction. In a real-robot experiment with a structured-light sensor, the adapted method achieves higher measured surface coverage with 60% fewer captures than nominal selection. The adapted method recovers a surface region that the nominal method repeatedly targets but fails to measure. These results suggest that robots can reuse pretrained predictive models while adapting their decisions to local operating conditions in data-scarce environments.
Read the full abstract

Machine Learning (Statistics)

Continuous Horizon Forecasting

Multi-horizon forecasting has been dominated by discrete forecast strategies that treat model outputs as fixed-vectors or recurrently create them. In this paper, we introduce the Continuous Horizon (CH) strategy that generalizes upon recursive, direct, and multi-step strategies by redefining forecasts as a continuous operator and the forecast horizon as a query variable rather than an index. We present the theoretical foundations for the CH strategy, a Universal Operator Approximation Theorem, and a continuous-horizon bias-variance decomposition that formally proves the estimation variance improvements of CH over the discrete strategies. We validate the CH strategy empirically through a simulation study, and on 16 large-scale datasets from the M-series benchmark, showing how CH achieves competitive RMSE accuracy, while uniquely enabling parallel inference, targeted horizon querying, and interpolation and extrapolation across unseen horizons.
Read the full abstract

All authors Malcolm Wolff, Elliott Zackrone, Souhaib Ben Taieb, Michael W. Mahoney, Kin G. Olivares

Robotics

Latent Barrier Steering: Hierarchical Safety for Generative Planning

We introduce Latent Barrier Steering (LBS), a hierarchical semantic-to-path framework for safe flow-matching-based generative planning. Existing control-barrier-function (CBF) safety filters for diffusion and flow-matching planners act locally on generated samples, either during sampling or after prediction. Such repair is effective for small violations, but can struggle when a new constraint blocks the whole selected behavior mode rather than merely perturbing it locally. LBS decouples behavior selection from path certification through two coordinated interventions: it first steers the generator's behavior latent toward rollouts with larger safety margin, then decodes the steered latent and applies a path-space CBF quadratic program (QP) only for residual violations. We show that latent steering increases the margin available to the path-space corrector, yielding finite-time recovery guarantees and reduced correction burden. Experiments across navigation and robot end-effector planning tasks show that LBS preserves safety while improving goal reaching and reducing correction burden compared with path-level safety filters.
Read the full abstract

Combinatorics

Unilluminable rooms: minimal examples, classifications and exceptions

A room with mirrored walls can contain two points that are mutually invisible: no light ray from one reaches the other. Tokarsky built the first polygonal examples by tiling a room with reflected copies of a triangle and making every tile vertex of certain kinds a corner of the room. We study the smallest rooms of this kind, which we call lattice rooms. Their interior tile vertices of the remaining kind, the sources, are invisible and pairwise mutually invisible. On the tiling by isosceles right triangles, a hole-free lattice room with a source has at least 12 sides, attained only by the pinwheel. On the square tiling the least number is 16, again attained by one room. A hole-free lattice room with two sources has at least 24 sides on the first tiling and 34 on the square tiling. Up to congruence there are 12 such rooms with 24 sides, 89 with 25 sides and 14 with 34 sides. One absorbing post, one partition, or one hole that touches the outer boundary at one point lowers the least number of sides for two sources to 18, 20 and 22 (21 if sides are counted as maximal straight segments). Two and three posts give 17 and 16 sides, and a hole that does not touch the outer boundary gives 26. We list all smallest rooms. Rooms assembled from n modules have exactly 12n sides and n sources, and a hole-free lattice room with three or more sources has at least 36 sides. We also give rooms in which some tile vertices need not be corners: 15 types of 27-sided rooms on the tiling by triangles with angles 30, 60 and 90 degrees, which are smallest in their class, a nonagon with three mutually invisible corners, and a 42-sided room in which a rotation through 180 degrees takes the place of a corner. In a ring of three trapezoids a light ray returns to a corner that is invisible in one trapezoid. Several classifications rely on exhaustive computer searches made finite by the bounds proved here.
Read the full abstract

Astrophysics of Galaxies

Intermediate-redshift slow rotators as the progenitors of the most massive galaxies at z $\sim$ 0

We study the resolved stellar kinematics and star formation histories, and their dependence on the environment, in a sample of 218 intermediate redshift (0.1 < z < 1) galaxies located in the CANDELS and COSMOS fields, observed with VLT/MUSE. The sample covers galaxies with stellar masses ranging M_star ~10^8 - 10^12 M_sun, including both star-forming and quiescent systems. For 113 galaxies, found in dense environments of galaxy groups, we derive the spin parameter ($\lambda_{R}$), while for the rest of the sample, we use data from previous papers in this series. Given that the surveys used in this work are flux-limited, the $\lambda_{R}$ distribution of our full sample lacks the bimodality observed in the local Universe, but the subsample of galaxies at z < 0.6 resembles the galaxy population in the nearby Universe, in terms of mass, sizes, and stellar population properties, more than galaxies at z > 0.6. We classified the galaxies into fast and slow rotators, finding that ~8% (17/218) of the galaxies are slow-rotating. The fraction of slow rotators does not change significantly down to z = 0, but we find that in the z ~ 0 sample, the spin parameter is, on average, higher than at higher redshifts. In our subsample at z > 0.6, we observed evidence for the existence of the kinematic morphology-density relation. The slow rotators in our sample mostly assemble their stellar mass in one or two main episodes of formation, and they finished assembling their mass via star formation by redshift of z $\sim$ 1. It is however, not clear if the loss of angular momentum can be associated with the last star formation episode. Based on their location in the mass - size diagram, the intermediate-redshift galaxies are consistent with being the progenitors of the most massive slow rotators observed at z~0, where the subsequent growth is facilitated through gas-poor (dry) mergers.
Read the full abstract

All authors C. Muñoz López, Davor Krajnović, B. Epinat, T. Contini, I. Pessa, J. Pharo, A. de Araujo Carvalho, A. R. Chougule

Data Structures and Algorithms

An Improved $(1+2e)$-Approximation Algorithm for Metric Data Placement

We study the metric data placement problem with site-capacity constraints and fixed installation costs, aiming to minimize the total cost of placing data (resources) and accessing their nearest copies. We develop a deterministic polynomial-time approximation algorithm based on a natural linear programming relaxation of the problem. Starting from an optimal fractional solution, we construct multiscale metric-ball partitions independently for each resource and connect overlapping balls at consecutive levels to form a flow network. The resource layers are then coupled through shared site-capacity arcs, allowing us to obtain an integral placement by minimizing a surrogate cost over a single min-cost circulation. By optimizing the choice of ball radii, we establish an approximation factor of $1+2e\approx 6.44$. To our knowledge, this is the best approximation guarantee for the metric data placement problem since its introduction by Baev and Rajaraman in 2001.
Read the full abstract

Computation and Language

Deletion Sensitivity Is a Property of the Model: A Three-Condition Measurement over 72 Valid Experimental Cells

Reasoning models write a trace at every turn of a multi-turn tool session, and later turns read it. Upstream providers hold different retention constraints, so a request forwarded across issuers arrives with its history deleted, and its task cost remains unmeasured. This paper reports a direct, blind, multi-seed measurement of full history deletion in multi-turn tool calling, through three rendering conditions, a calibrated set of 72 valid experimental cells, a multi-seed protocol, and the official checker. The same deletion separates the models by direction and by magnitude, one class losing tens of points, another gaining, and a third unchanged, with the endpoint ordering reproduced across seeds and task subsets modulating the magnitude and, near zero, the sign. A re-dosed model preserves its success across the full deletion range, so the reading belongs to the model rather than the dose. Deletion sensitivity is a model-level property under an operational definition, measured under a fixed task family, scorer, and protocol, and a retrained model is a new model that a two-cell screen, one retention run and one deletion run, re-reads for the cost of two runs. Deletion sensitivity is thus a measurable property of the model, and the reading must enter routing as prior information rather than persist as an invisible default cost.
Read the full abstract

All authors Yiran Li, Hanzhang Jia, Liheng Zeng, Zhiyu Yang, Yi Gao