Instrumentation and Detectors
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Sunday 8 November
The PTOLEMY project aims at measuring the neutrino mass and, in the future, the Cosmic Neutrino Background from Tritium endpoint measurement. The Radio-Frequency (RF) Tracker is a sub-detector with the goal of measuring in real time the electrons by exploiting the Cyclotron Radiation Emission Spectroscopy. In this work, an overview of the PTOLEMY project is given together with the design, simulation, construction and commissioning of the RF Tracker prototype: the Electron Trap. The measurements and data analysis of the Electron Trap are presented, establishing the primary path for the development of the RF Tracker.
Group Theory
Uniform Hyperbolicity of Automorphism Orbit Graphs in Rank Two
We study automorphism orbit graphs of oriented conjugacy classes in the free group of rank two, with every elementary Whitehead move assigned unit cost. Collapsing signed permutations of the basis turns each orbit component into a tree of edges and triangles. We obtain this description by combining the known rank-two conjugacy-class stabilizer structure with the Bass-Serre tree of the rank-two outer automorphism group. The simplified graph and all its vertex-induced subgraphs have optimal thin-triangle constant zero for tree components and 3/4 otherwise. Every whole orbit component of the original graph has thin-triangle constant at most 2, while every component of every vertex-induced subgraph has constant at most 5/2. Both bounds are sharp in the primitive orbit and apply to the full geometric realization, including edge-interior triangle corners. The hereditary upper estimate also follows from known chordality theory; the whole-component improvement uses the geometry of adjacent permutation fibres. These results complement earlier fixed-length and minimum-length orbit classifications.
Read the full abstract
Software Engineering
Just AI Governance and the Forbidden-Boundary Principle
Artificial intelligence (AI) increasingly shapes consequential decisions, yet predictive accuracy is not a sufficient justification for delegated authority. This conceptual study develops an integrated framework for assessing just AI across normative, data, model, and governance layers. Drawing on plural theories of justice, sociotechnical fairness research, and publicly available governance standards, it maps six justice requirements to observable evidence, decision rights, and remedies. It proposes a two-stage evaluation rule: exclude uses that fail independently justified minimum rights constraints, then compare eligible alternatives across multiple performance and governance criteria. As an explicitly limited normative analogy, Genesis chapter 2 motivates a Forbidden-Boundary Principle (FBP), under which an AI agent must not execute externally consequential actions outside pre-authorized boundaries. The resulting First-Principle Safe-Stop Protocol (FPSSP) comprises an independently governed policy monitor, a mandatory execution gate, a restricted or alternative safe-state controller, immutable-oriented audit trails, and accountable human authorization for recovery. We distinguish this normative and traceability contribution from existing Simplex and runtime assurance architectures; we do not claim a novel safety switch or formal proof of universal ethical compliance. A policy schema, threat model, test matrix, outcome definitions, and a prospective comparison design make the proposal falsifiable. No FPSSP implementation, experimental dataset, numerical improvement, or verified latency bound is reported. The framework supports future evaluation rather than constituting evidence of operational safety.
Read the full abstract
Artificial Intelligence
Measuring and Attributing Evidence Sensitivity in Medical Vision-Language Models via Symmetric Counterfactual Controls
Medical vision-language models (VLMs) asked closed clinical questions ("Is there a polyp?") can score well by relying on the question rather than the image, and current hallucination benchmarks cannot separate true visual discrimination from editing shortcuts or shifts in the answer threshold. We introduce a Counterfactual Evidence Framework that pairs each lesion erasure with a shape- and tool-matched control edit elsewhere in the image, adds same-site ring controls, restored and partial erasures, degraded views and natural negatives, tests on held-out inpainting tools, separates discrimination (the Evidence Sensitivity Gap, ESG, and AUROC) from the answer threshold, and checks the labels with blinded physician readers and hidden-state probes. Across 507 main-test cases (1,115 views), nine frontier VLMs claim 41.1% to 61.6% of erased lesions (ESG 0.269 to 0.470), whereas MedScope-4B, a Qwen3.5-4B agent trained on automatically built counterfactual trajectories, claims 9.6% (ESG 0.758; p <= 0.001 in all 27 primary comparisons). Physician readers support automated erasure in colonoscopy, where frontier models still claim 45% of erased lesions against 1% for MedScope-4B, but call a finding present on 54.7% of pooled erased reads across all sources, which bounds what unverified erasure can certify. Across 12 pre-registered training or system arms and 270 decision policies screened under a rule fixed in advance, every apparent utility gain comes from shifting the threshold rather than improving held-out discrimination, and probes point to perception as the remaining limit. We also report a 0/48 pneumothorax failure on an external cohort; retraining with pneumothorax patients in every stage finds 53% of pneumothoraxes on new patients without changing the main-test results.
Read the full abstract
Machine Learning
The Shape of a Program: Path Signatures for Trace-to-Program Induction
Synthesizing programs from execution traces is a fundamental challenge in algorithmic interpretability and the reverse engineering of complex systems. We bring a rough-path-theoretic perspective to program induction, treating execution traces as high-dimensional rough paths and encoding them with the path signature transform, a tool from stochastic analysis whose expectation provably characterizes the distribution of traces a program induces. On three program-induction domains, Karel, List, and RASP, our log-signature trace encoder outperforms LSTM, Transformer, and Fourier baselines in almost every setting, with its largest margins when training data is scarce. We further observe that the trace-to-program pipeline is naturally contractive: it often produces valid programs that are shorter than the ground truth. We turn this into a bootstrapping procedure that exposes redundancy in the ground-truth supervision without any external oracle, which we demonstrate on redundancy-injected versions of our datasets. Finally, the log-signature encoder induces a markedly more interpretable latent geometry than the baselines, in which the nearest neighbors of every state lie on its own execution trace.
Read the full abstract
Computer Vision
G-STAR: Group-Scaled Token Capacity for Deeply Compressed Autoregressive Generation
Discrete autoregressive models offer a scalable framework for image generation. However, they face a fundamental trade-off: long token sequences preserve representation quality but slow generation, while aggressive spatial compression reduces representation capacity. Based on product quantization, we show that, under a fixed nominal information budget, group size should scale quadratically with the per-axis spatial compression factor. Building on this principle, we propose GSTAR, a spatial–group co-scaling framework that enables \(32\times\) and even effective \(64\times\) compression while preserving strong reconstruction, understanding, and generation performance. Since larger groups introduce intra-group dependencies that must be modeled efficiently, we analyze their impact on inference latency and generation quality and introduce a semi-autoregressive decoder that combines parallel feature computation with lightweight causal modeling. With these designs, our highly compressed models perform competitively with leading models across five representative text-to-image benchmarks (GenEval, DPG-Bench, PRISM, LongText-Bench, and Qwen-Image-Bench). At \(1024\times1024\) resolution, our model delivers a \(20\times\) speedup over prior discrete autoregressive models, generating each image in 2.5 seconds on a single NVIDIA H200. Together, these results show that highly compressed discrete representations can enable high-quality autoregressive generation at low latency.
Read the full abstract
Machine Learning
Branch-and-Commit for Frozen RL Agents: When and Where to Spend Test-Time Compute
Reinforcement learning agents are often deployed with policies that cannot be retrained or modified. In digital environments whose states can be copied and resumed, an operator has another option: sample several continuations from the current state, evaluate them, and continue from the endpoint of the selected branch. We call this procedure branch-and-commit. We study when it improves return, what produces the improvement, and where its computational budget should be spent. Across four environments, two frozen policies, five continuation verifiers, and a model-free selection rule, we find that the benefit of branching depends on dispersion: the difference between the best candidate return and the mean candidate return. Branching provides no benefit in low-dispersion settings, while increasing return by 12.8% in our main Craftax-Medium setting. Most of this gain comes not from selecting better actions, but from selecting among different stochastic realizations of the same action. Several competent verifiers achieve similar episode returns, suggesting that verifier quality is not the main constraint at our operating budget. Instead, computation is better allocated to states where useful dispersion is likely. On held-out seeds, gating branching using predicted dispersion retains 69% of the improvement from uniform branching while using 48% as many branching calls. These results characterize when branch-and-commit is useful and clarify what its gains represent.
Read the full abstract
Multimedia
YTShorts-92K: An Interpretable Multimodal Dataset of YouTube Shorts
Short-form video has become one of the dominant formats of online content, yet large-scale datasets enabling the analysis of what makes these videos successful remain limited. Existing resources often focus on post-publication popularity signals or rely on non-interpretable representations such as high-dimensional embeddings. Here, we introduce YTShorts-92K, a dataset of 92,654 YouTube Shorts collected from 345 creators and spanning 12 manually defined content categories. Each video is enriched with an interpretable multimodal feature set covering metadata, textual information, visual characteristics from frames and thumbnails, and cross-modal consistency measures. We describe the collection methodology, data cleaning process, and feature extraction pipeline, followed by descriptive analyses of the diversity of creators, topics, and extracted features. This dataset offers a reusable resource for studying short-form video content, online attention, and multimodal content analysis.
Read the full abstract
Machine Learning
One Table, Many Queries: Benchmarking Generalist Tabular Models Across Generation, Imputation, and Prediction
Generation, imputation, and prediction query different aspects of the same tabular distribution, but are usually evaluated separately. This leaves unclear which capabilities they share, how their training objectives interact, and whether one model can perform all three well. We introduce the One Table Benchmark (\ours{}), spanning 56 datasets, 46 registered methods, and 20 capability axes under aligned data splits and explicit task contracts. We compare task ranks on aligned support, train five backbones under all seven non-empty objective combinations on five datasets with three seeds, and contrast fixed-model profiles with specialist portfolios. Rank comparisons connect imputation and prediction through numerical recovery and dependence preservation, and link generation to both through predictive utility; distributional fidelity follows different, sometimes opposing, rankings. Objective interventions show the largest pooled gains from including the recipient's own objective. Generation has the broadest coverage of positive pooled auxiliary effects, while prediction shows potential to improve imputation and synthetic-data utility, with responses that vary across backbones. Cross-addition effects reveal that objective benefits need not add and depend on the objectives already present. Fixed-model comparisons expose corresponding utility–fidelity trade-offs: some generalists surpass specialist portfolios on utility, while fidelity continues to favor specialization. Frozen foundation models show strong prediction and numerical point imputation, together with promising synthetic-data utility, highlighting their potential across all three tasks. These findings motivate unified tabular models that share useful conditional information while improving categorical reconstruction and faithful joint sampling.
Read the full abstract
Computer Vision
PhysMoE: Fast Spatial Grounding with Single-Token Coordinate Representation in Modality-Decoupled VLMs
Embodied Vision-Language Models (VLMs) are backbones for robotic planning and manipulation. However, for coordinate prediction in grounding tasks, existing architectures suffer from two distinct bottlenecks: discrete tokenization disrupts spatial continuity with high decoding latency, while standard point regression fails on multi-contact affordances and irregularly shaped targets. To resolve these challenges, we propose PhysMoE, introducing three core contributions: (1) sharing lower layers for multimodal context while routing upper layers to specialized spatial (pixel) and linguistic (text) experts by physical modality, eliminating gradient interference; (2) encoding each coordinate as a single token by reusing the visual encoder's learned positional embeddings, injecting 2D spatial priors to natively align coordinates with visual features and ease training; and (3) differentiable Signed Distance Field (SDF) supervision on pointing, accommodating affordances and irregular geometries without costly mask decoders. Compared to the base model, PhysMoE achieves a 15.7% average gain and an 8.31x throughput speedup across 9 coordinate-based grounding benchmarks (trajectory, affordance, and localization), while preserving multimodal understanding and reasoning across 11 text-based embodied benchmarks.
Read the full abstract
Atmospheric and Oceanic Physics
Direct Multispectral Satellite Nowcasting of Gridded Surface Solar Radiation
Surface shortwave radiation (SWR) can change within minutes as clouds evolve, but operational forecasts often trade spatial detail and update frequency for forecast range. Here, we present a numerical-weather-prediction-independent, observation-to-forecast framework that converts sequences of Himawari-9 Level-1 multispectral imagery and solar geometry directly into 24 gridded SWR fields at 10-min intervals for the next 4 h. A controlled six-model experiment in eastern China separates the effects of satellite information from those of temporal self-attention. In the 0-1 h window, the full-multispectral attention model achieves a mean absolute error of 37.9 W/m2 and a root mean square error of 63.7 W/m2, reductions of 21.0% and 19.5% relative to a historical-SWR baseline. Improvements extend across nine subregions and include weaker spatial bias and better structural consistency. Temporal attention does not improve forecasts from historical SWR alone, but becomes effective when supplied with multispectral cloud and thermodynamic signals. Direct multispectral observations therefore provide the precursor information required for high-frequency, spatially resolved SWR nowcasting.
Read the full abstract
Machine Learning
Linear Probes to Bayesian Manifolds: Geometry of Arithmetic in Large Language Models
Linear probes are a standard method to study geometric structures, such as helices and circles, from the activations of large language models (LLMs), and such structure is often taken as evidence that the model computes with it. Probe accuracy alone cannot establish this, because it conflates four separate claims. These are (1) that the concept is decodable from the activations, (2) that they form a specific shape, (3) that the shape belongs to the probed concept rather than to a related concept the model also represents, and (4) that the output of the model depends on the shape. We call these claims decodability, shape, ownership, and causal use, and we test them separately with a four-stage Bayesian pipeline. Stage 2 compares candidate shapes and selects the best one only when its evidence passes a predefined threshold. Stage 3 removes a set of related concepts, specified in advance, and tests whether the shape is owned or inherited. Stage 4 intervenes on the shape to test whether the output of the model depends on it. We apply the pipeline to two-digit addition and multiplication in GPT-J 6B, Llama 3.1 8B, and Pythia 6.9B. Nearly every concept is decodable, but the decoded directions overlap substantially across concepts. Every strongly supported comparison selects a ribbon, which is a helix whose radius changes with the value, or a torus. Of 1,330 cells with an ownership verdict, 19% of shapes are owned and 81% are inherited. Owned and inherited shapes move the output about equally often. Intervening on the shape retains approximately 61% of the effect of intervening on the full probe while using at most five dimensions, against a median of nine for the probe. Our results indicate that a probe verdict alone overstates what a model computes with, and we propose that probe results be reported as verdicts on all four claims rather than as a single accuracy.
Read the full abstract
Computer Vision
TIB: Sample-wise Tempered Information Bottleneck for Multimodal Attribution beyond Alignment Assumption
Multimodal (e.g. vision-language, in this work) attribution aims to interpret the models without access to ground-truth explanatory supervision. Existing multimodal attribution methods typically rely on paired modalities as proxy supervision, implicitly assuming that image-text pairs are semantically aligned. Information bottleneck (IB)-based attribution methods follow this general paradigm and instantiate it through a global sufficiency-compression objective. In realistic multimodal datasets, however, this assumption is frequently violated due to partial semantic misalignment and unreliable cross-modal correspondences. Through theoretical analysis, we show that, under such misalignment, information-bottleneck-based attribution objectives exhibit a structural limitation: by assuming equal reliability across all proxy supervision signals, they induce posterior over-contraction and forced explanations. To address this issue, we propose TIB, a sample-wise Tempered Information Bottleneck framework that adaptively modulates attribution strength according to cross-modal reliability, without assuming semantic alignment. Extensive experiments on controlled benchmarks and large-scale datasets show the effectiveness of the proposed approach, while also validating the identified failure mode.
Read the full abstract
Numerical Analysis
Geometry-Dependent Dimension Jumps in Planar $C^r$ Splines at Degree $2r+1$
We study geometric dependence of the dimensions of planar spline spaces $S_{2r+1}^r$ through families with a single moving interior vertex and determine their dimensions throughout the admissible parameter domains. For $r=1$, we extend the Wu–Deng one-parameter classification to the interior of the central quadrilateral. The dimension increases by one at exactly two isolated points, one due to a local direction-count change and the other to a global compatibility defect. For $r=2$, subdividing one triangle of the Yuan–Stillman configuration gives a twelve-triangle family with fixed local direction counts; its dimension increases by one precisely on an open line segment. For every $r\ge3$, we construct a family with $2r+12$ triangles and fixed local direction counts, proving geometric dependence at degree $2r+1$. The moving vertex lies in an open triangle, and the dimension exceeds Schumaker's lower bound by one exactly on its symmetry axis. The lower collinear vertices may have arbitrary strictly ordered heights. The low-order proofs use exact cofactor elimination. For higher orders, weighted interpolation and block elimination bound the global defect by one, and explicit extension and obstruction arguments determine the exceptional set.
Read the full abstract
Computation and Language
Capacity Allocation at the Source: Sparse Target Optimization for LLM Knowledge Editing
Knowledge editing has emerged as a promising approach for efficiently updating embedded knowledge in large language models (LLMs). It first computes an ideal target hidden state that steers the LLM toward the new output, and then edits the model parameters so that the original hidden state is mapped to this target. While most existing methods focus on the second stage of parameter editing in preserving other knowledge, we find that the upstream construction of the target hidden state also plays an important role in allocating the model's limited capacity space. We first establish a theoretical connection between interference in the parameter space and interference between target hidden states, showing that the two are positively correlated. Building on this insight, we propose a simple yet effective $\ell_1$ regularization method that prunes away unnecessary “tentacles” of each knowledge vector, retaining only the essential components. With fewer extraneous tentacles, each edit affects less other knowledge, enabling better coexistence. Extensive experiments across multiple LLMs and benchmark datasets verify the effectiveness of our method. Code: https://github.com/wintersdragon-c/STO.
Read the full abstract
Nuclear Experiment
Measurement of the $^{169}$Tm$(\gamma,n)$$^{168}$Tm Cross Section and Implications for $\gamma$-Process Calculations
The $^{169}\mathrm{Tm}(\gamma,n)^{168}\mathrm{Tm}$ cross section has been measured at the Shanghai Laser Electron Gamma Source (SLEGS) using quasi-monoenergetic laser-Compton-scattering $\gamma$ rays over the energy range from the neutron threshold to 14.61~MeV. The photon flux and energy distribution were determined with a LaBr$3$(Ce) spectrometer, and monoenergetic cross sections were obtained using an iterative unfolding method. Below 12~MeV, the present cross sections agree well with the NewSUBARU data and the IAEA-2019 evaluations. However, near the maximum of the $(\gamma,n)$ excitation function, the present cross sections are significantly higher than both the NewSUBARU data and the IAEA-2019 evaluated values. The $E1$ $\gamma$-ray strength functions were derived by fitting the experimental cross sections measured at SLEGS and NewSUBARU. The stellar reaction rates for $^{169}\mathrm{Tm}(\gamma,n)$ and $^{168}\mathrm{Tm}(n,\gamma)$ were calculated by incorporating these constrained $\gamma$-ray strength functions into the TALYS code. The calculated $^{169}\mathrm{Tm}(\gamma,n)$ rates were adopted in a full $p$-process reaction network for the shock-heated O/Ne-rich layers of a $20 M\odot$ core-collapse supernova model to perform a sensitivity analysis. The updated rates have a negligible effect on the final $p$-nucleus abundances.
Read the full abstract
Quantum Physics
Second-level termination of the PPT entanglement-cost hierarchy for two qutrits
We prove that for arbitrary two-qutrit quantum states, the semidefinite programming lower-bound hierarchy for the PPT entanglement cost terminates at the second level. Therefore, under operations preserving positive partial transpose, its zero-error asymptotic entanglement cost can be obtained by a semidefinite program involving only three Hermitian matrix variables. The key to the proof is to shorten the feasible dual chain by one level; to this end, we combine finite certificates verified by exact rational arithmetic with analytic estimates near the boundary of the parameter domain. This result shows that a unified terminating level exists for two-qutrit systems, whereas no unified terminating level exists for all dimensions.
Read the full abstract
Artificial Intelligence
ZeroPoint: Lightweight Energy-Based Decision Model
Energy-based models (EBMs) assign a scalar energy to every joint value of the observed and predicted variables. In decision models, the prediction is chosen from a finite list of options provided with each input. Inference then reduces to comparing the energies of the options, and training lowers the energy of the correct option relative to the others. A complete decision must also account for the possibility that no option is acceptable. However, the energies learned by a standard EBM define only relative compatibility: adding the same input-dependent quantity to all energies leaves the predictions, the Gibbs distribution, and the usual training losses unchanged. The model therefore has no learned reference level against which to decide whether any option fits. We propose ZeroPoint, an EBM that makes the empty answer a state of the model with energy zero for every input. This fixed state provides an explicit reference for the energies of all listed options, whose non-negative depths measure their compatibility with the input relative to the empty answer. The options and the empty answer share one distribution, in which an option at zero depth receives no probability. The decision to select an option or to abstain is thus part of ordinary inference, trained with the negative log-likelihood loss, rather than a separate rejection rule. The energy function is a small multilingual encoder that reads the input together with all the options and is invariant to their order. Ablations show that most of the benefit comes from training with examples in which no option fits. With the same encoder and data, the fixed zero separates inputs with and without an acceptable option slightly better than a learned rejection class and clearly better than a thresholded softmax, while leaving the choice among options unchanged.
Read the full abstract
Software Engineering
Specializing Meta Multi-Objectivization as Baseline for Software Engineering Optimization
Given the discrete nature of many software engineering (SE) optimization problems, trapping at local optima remains a crucial issue in the optimizer designs. Meta Multi-Objectivization (MMO) is an optimization model for overcoming local optima by introducing an extra, auxiliary objective. Yet, MMO is originally designed only for configuration tuning, leaving its suitability for other SE problems in doubt, particularly regarding the choice of the auxiliary objective. In this paper, we study the possibility and resolution for specializing MMO as a general baseline across a range of SE optimization problems, making it a fundamental counterpart for future SE optimizers that seek to mitigate the issue of local optima. We do so by providing both theoretical remit and empirical evidence: we analytically show how the choice of the auxiliary objective is crucial while demonstrating the effectiveness of MMO under nine SE optimization problems comprising 212 cases when the right one can be chosen, i.e., MMO is able to outperform the other state-of-the-art and general-purpose optimizers in 79% and 94% of the cases, respectively. Drawing on the theoretical understanding and empirical findings, we then propose L2RAC, a tool that automatically recommends the best auxiliary objective choice for a given SE problem/case, hence helping to specialize MMO therein. Through extensive experiments, we reveal that, in general, L2RAC can more accurately rank the choice of auxiliary objective for MMO under a case than other guesses by up to 60%, saving as much as 99% of the cost that would otherwise be required to conduct trial-and-error when using MMO.
Read the full abstract
Machine Learning
Multimodal Graph Learning: A Survey of Models, Benchmarks, and Frontiers
Multimodal attributed graphs (MAGs), in which nodes carry text, images, and other modalities while edges encode relations among entities, underlie e-commerce, social media, content platforms, and creative applications. Learning on MAGs has developed along two complementary lines: graph-centric methods redesign how modalities are fused, propagated, and connected within message passing, while foundation-model-centric methods pretrain transferable multimodal graph encoders or bring graph context into multimodal large language models for prediction and generation. In parallel, several benchmarks have appeared with different datasets and evaluation protocols. This survey takes an algorithm-centric and evaluation-grounded view of the field and makes three contributions. Modeling: we introduce a unified message-passing formulation and organize methods into five integration levels (feature, propagation, topology, learning, and foundation model), covering discriminative and graph-conditioned generative models; within this formulation we derive representative methods as special cases, compare their assumptions, expressiveness, and cost, and identify cross-modal sensitivity as a property that separates otherwise similar architectures. Datasets and evaluation: we review recent benchmarks, summarize their datasets and metrics, define model-independent descriptors (label homophily, average degree, clustering, and edge provenance), and map each evaluation protocol to the baselines it uses. Frontiers: we connect multimodal graph learning to AI agents, covering the multimodal graphs that agents perceive and remember and graph engineering, the practice of organizing agent workflows as explicit graphs; we then discuss open problems, concrete efforts for the community, and research directions, including when graph structure helps in the presence of strong multimodal foundation models.
Read the full abstract
Populations and Evolution
A Markovian SuASR epidemic framework with stochastic external factors
Epidemic waves are often shaped by uncertain external drivers, such as environmental variability or behavioral and policy changes, that can randomly amplify or dampen transmission. Capturing this randomness requires models whose dynamics evolve under time-varying, stochastic influences. In this paper, we introduce a stochastic Susceptible-Asymptomatic-Symptomatic-Removed (SuASR) framework that embeds random external factors directly into the epidemic mechanism through an extended continuous-time Markov chain (CTMC). The proposed framework is flexible enough to represent epidemic dynamics both with reinfection and with infection-induced immunity, depending on the epidemiological setting under study. Unlike standard Markovian epidemic models, our formulation explicitly tracks two concurrent infectious sources (asymptomatic and symptomatic), yielding coupled infection pathways and substantially richer transient behavior. A key contribution is analytical tractability, where we derive computable expressions for the joint distribution of asymptomatic and symptomatic infection-event counts and develop tools tailored to extreme-outbreak risk, including the distribution of the maximum infectious burden. These results enable probabilistic quantification of epidemic severity beyond mean trajectories, providing direct estimates of the likelihood of rare but consequential surges. We validate the model through extensive sensitivity analysis and an application to the H1N1 influenza epidemic, illustrating how stochastic external influences reshape outbreak profiles, total infections from both sources, and the probability of extreme epidemic scenarios
Read the full abstract
Machine Learning
Third Eye: Beyond Immediate Gains in Recursive Self-Improvement
Does a useful self-update remain useful after the model updates again? Third Eye separates immediate performance from the outcome of a further native update. We evaluate five 3-8B backbones on math and code: 64 fresh runs, three candidates per run, and three continuations per candidate. Selection and final scoring use disjoint items. In the balanced Qwen3-4B/Llama-3.2-3B comparison, immediate-selected utility is near zero, but continuation reduces it by 1.78 percentage points (simultaneous 95% interval [0.68, 2.88]). Candidate-averaged immediate utility is already negative: selection partly accounts for the apparently neutral first step. Continuation also lowers utility for SmolLM3-3B and Llama-3.1-8B; Qwen3-8B remains inconclusive. Target and out-of-distribution losses dominate retention changes. Matched recipe and recipient-donor controls identify neither a reliable mitigation nor a single mechanism. The results support evaluating persistent self-updates as adaptation transitions, with independent scoring and a no-update baseline, rather than interpreting immediate performance as their full value.
Read the full abstract
General Economics
Global Trends and Heterogeneity in Modified Adjusted Net Savings
Modified adjusted net savings is a derived weak sustainability indicator obtained by subtracting education expenditure from adjusted net savings (ANS). Because ANS equals net national savings plus education expenditure minus natural-resource depletion and carbon dioxide damage, MANS isolates the savings and depletion components and is useful when education expenditure would otherwise dominate cross-country comparisons. This study documents the global distribution, long-run trends, and cross-country heterogeneity of MANS excluding particulate emission damage over 1990-2021. The accounting base is the World Bank World Development Indicators series NY.ADJ.SVNX.CD, and MANS is constructed by subtracting education expenditure from this series. We combine global maps, aggregate time series, country-level heatmaps, distributional statistics, and income group comparisons. The results show that global MANS has increased substantially in aggregate, but its evolution as a share of nominal GDP has been more volatile and uneven across countries. Per capita MANS and global MANS totals remain concentrated in high-income and large economies, whereas low-income and lower-middle-income groups display persistently lower and more volatile outcomes. Country rankings are highly persistent, and some populous countries exhibit pronounced fluctuations associated with natural-resource dependence and structural change. The study provides a reproducible descriptive benchmark for monitoring sustainable development and highlights the need to interpret MANS in current US dollars alongside income-group and country-specific contexts. The findings are relevant for macroeconomic policy, natural-capital accounting, and the design of sustainability-oriented investment strategies.
Read the full abstract
General Physics
Dust matter cosmology in modified symmetric teleparallel $f(Q,T)$ gravity
We investigate the nonlinear $f(Q,T)$ gravity model, $f(Q,T) = \alpha Q^n + \beta T + \gamma$, which represents a nontrivial coupling between geometry and matter in the context of a dust-dominated Universe. The model parameters $(H_0, \alpha, \beta, \gamma, n)$ are constrained using recent cosmological observations, including Hubble parameter measurements $H(z)$, the Pantheon+SH0ES dataset, and DESI BAO data. The analysis shows that the Hubble constant exhibits dataset-dependent variations, reflecting the persisting Hubble tension, while $\alpha$ and $\gamma$ are tightly constrained and $\beta$ remains weakly determined. The nonlinear exponent $n$ introduces an additional degree of freedom that enhances the model's flexibility in reproducing observational trends. The cosmographic diagnostics, including the deceleration, jerk, snap, and lerk parameters, indicate a Universe transitioning from early deceleration to late-time acceleration, with present values closely aligned with $\Lambda$CDM predictions and far-future behavior approaching a de Sitter phase. These results demonstrate that the nonlinear $f(Q,T)$ model provides a simple and viable framework to describe late-time cosmic acceleration, accommodate observational tensions, and reproduce $\Lambda$CDM-like dynamics with subtle deviations.
Read the full abstract
Programming Languages
Parametric Proof-Carrying Compilation for Quantum Error-Correction
We present PC-QEC, a proof-carrying compilation framework for quantum error correction parameterized by both code distance and syndrome-extraction scheme. It covers all four schemes—Standard, Flag, Shor and Knill—with circuit-level distance guarantees for every legal code distance in each proved family under the declared error and readout contracts. The circuit-level model permits single-wire Pauli errors at every operation's qubit operands, capturing data errors, readout errors and ancilla errors that propagate into multi-qubit hook errors. For each family and scheme, supplied code-algebra and invariant proofs combine with proved scheme-specific propagation lemmas into a finite symbolic certificate whose size is independent of code distance. We demonstrate family-wide guarantees for five code families across all four schemes. For rotated Surface code, one Standard Scheme certificate proves a distance lower bound $d$ for every odd $d\ge3$ and checks in 214\,s on average. Lean 4 mechanizes the metatheory and exact circuit-level distances for canonical rotated Surface, unrotated Surface and Toric code families.
Read the full abstract
Analysis of PDEs
Uniform Resolvent Estimates for the Dunkl Laplacian
Let \(\Lk=-\Dk\) denote the nonnegative Dunkl Laplacian associated with a finite reflection group and a nonnegative multiplicity function \(\kappa\), and let $ N=d+2\gk\geq 3 $ be the corresponding homogeneous dimension. We establish uniform \(L_\kappa^p(\mathbb{R}^d)\)-\(L_\kappa^q(\mathbb{R}^d)\) resolvent estimates for \(\Lk\) whenever \[ \frac{2}{N+1} \leq \frac{1}{p}-\frac{1}{q} \leq \frac{2}{N}, \quad \frac{2N}{N+3}<p<\frac{2N}{N+1}, \quad \frac{2N}{N-1}<q<\frac{2N}{N-3}, \] where the last upper bound is interpreted as \(q<\infty\) when \(N=3\). Remarkably, this admissible range coincides with the corresponding range for the Euclidean Laplacian \(-\Delta\) in dimension \(N\).
Read the full abstract
Robotics
CAFE: Closed-Loop Autonomous Task Formulation and Execution without Instance-Specific Instructions
Scene-driven robot autonomy requires selecting tasks from observations as well as executing them. We study scene-driven task formulation and execution under predefined cleaning, safety, and storage/organization categories, without instance-specific goals or action sequences. CAFE maintains online memory of objects, functional zones, and execution states. A rule-guided Trigger-LLM admits formulation from accumulated evidence; Draft, Complete, and Audit propose goals, resolve execution details, and check dependencies. Outcomes update memory for correction and subsequent discovery. Evaluation in 12 AI2-THOR scenes compares three backbones and task-complexity tiers using coverage, constraint, and execution metrics. Tri-LLM records higher GCR scores than the evaluated baselines across three backbones and task-complexity tiers. Controls examine completion, auditing, zones, feedback, triggering, and prompt examples. LimoCobot traces illustrate storage, cleanup, and recovery from a failed grasp. The study examines how shared scene state connects task discovery to execution and outcome-driven revision.
Read the full abstract
Hardware Architecture
ICLayout-Bench: Benchmarking LLM Agents for Custom Circuit Layout Design
Large language model (LLM) agents are being explored for circuit design, but their ability to produce high-quality custom layouts has not been extensively evaluated. ICLayout-Bench is a benchmark in which agents use electronic design automation (EDA) tools to construct and optimize GDSII layouts from transistor-level circuit netlists. These layouts are evaluated using physical verification, parasitic extraction, and post-layout simulation. Layouts that pass the physical and functional checks are scored on electrical performance and area. The benchmark includes 165 tasks across 13 process design kits (PDKs), covering technology nodes from 3 nm to 180 nm. The tasks span analog, mixed-signal, and digital circuits in open-source and commercial EDA flows. The experiments evaluate 22 configurations spanning seven models on 21 public and 19 private tasks. Mean layout scores range from 4.22 to 80.26 on public tasks and from 1.43 to 67.01 on private tasks, on a 100-point scale. Higher reasoning effort does not consistently improve mean scores, and configurations with comparable scores can incur substantially different model costs.
Read the full abstract
Computer Vision
Evidence-Grounded Iterative Reasoning for the Diagnosis of Diffuse Cystic Lung Diseases
Overlapping imaging manifestations complicate the differential diagnosis of Diffuse Cystic Lung Diseases (DCLDs), while limited annotated data constrain supervised model development. Multimodal Large Language Models (MLLMs) can integrate imaging and clinical information, but their diagnostic reasoning remains limited by incomplete volumetric perception and insufficient grounding in disease-specific evidence. We propose DCLDs-EIR, an evidence-grounded iterative reasoning framework coupling diagnostic hypothesis refinement with multimodal evidence acquisition and verification. Multiview renderings of predicted lung and cyst segmentations characterize global cyst distributions and initialize the differential diagnosis. Competing hypotheses guide axial CT inspection to identify local findings that, together with clinical context, inform retrieval from a disease-specific knowledge graph. Accumulated patient observations and retrieved knowledge are jointly assessed for diagnostic support and refutation to revise the diagnostic state. Updated hypotheses and unresolved distinctions guide subsequent inspection, forming a feedback loop between evidence acquisition and hypothesis verification. Multicenter evaluation demonstrates that DCLDs-EIR consistently improves diagnostic accuracy over baselines and enables traceable differential diagnosis through explicit links between diagnostic hypotheses, patient observations, and disease-specific knowledge.
Read the full abstract
Machine Learning
MADSNet: A Multi-Scale Axial-Attentive Dense Spiking Network for Spike Camera Image Reconstruction
Spike cameras encode dynamic scenes into high-temporal-resolution binary spike streams for efficient high-speed vision. Although spiking neural networks (SNNs) can process such signals with sparse and accumulate-only computation, existing SNN-based reconstruction methods are typically shallow and still lag behind ANN-based models in reconstruction fidelity. This paper proposes MADSNet, a Multi-scale Axial-attentive Dense Spiking Network for spike camera image reconstruction. MADSNet adopts a hierarchical spiking encoder–decoder, where a multi-scale spike stem extracts early features, axial spike attention captures long-range dependencies, a dense spiking bottleneck enhances feature reuse, and a spike-to-continuous decoder reconstructs continuous intensity images from spike-domain representations. Experiments on gray-scale and color spike reconstruction show that MADSNet outperforms SSIR by 4.3 dB PSNR and Spk2ImgNet by 1.94 dB on SREDS. Compared with Spk2ImgNet, it achieves about 15$\times$ lower latency and uses only about 1/40 of the estimated energy, demonstrating that carefully designed SNN architectures can improve reconstruction quality while preserving computational efficiency.