Peer Revenue Now accepting positions

Papers

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

All of arXiv

Friday 6 November

Information Theory

The Schur products of some nonbinary narrow-sense antiprimitive BCH codes

In recent years, the Schur products of linear codes have been extensively studied due to their favorable applications in cryptography. In this paper, we aim to study the parameters of the Schur products of some nonbinary narrow-sense antiprimitive BCH codes. Let $\mathcal{C}_1$ and $\mathcal{C}_2$ be two narrow-sense antiprimitive BCH codes with designed distances $\delta_{1}$ and $\delta_{2}$, respectively, where $2\le \delta_{1},\delta_{2}\le n$. Let $\mathcal{C}_1\star\mathcal{C}_2$ be the Schur product of $\mathcal{C}_1$ and $\mathcal{C}_2$. We present some necessary and sufficient conditions for $\mathcal{C}_1\star\mathcal{C}_2\ne\mathbb{F}_q^n$ by applying some constraints to $\delta_{1}$ and $\delta_{2}$. Furthermore, we determine the dimensions of these products and lower bounds on their minimum distance. Several linear codes with best known parameters are found.
Read the full abstract

All authors Haojie Xu, Xia Wu, Wei Lu, Xiwang Cao

Complex Variables

Centralizers of elliptic univalent self-maps and the embeddability problem in the unit disc

Let $\phi$ be an elliptic univalent self-map of the unit disc in the complex plane that is not an automorphism, with multiplier $0<|\lambda|<1$ at its Denjoy–Wolff point. The simultaneous Schröder linearization of maps commuting with $\phi$ is classical. Using this linearization, we identify its univalent centralizer with the multiplicative semigroup of scalars that preserve the Koenigs domain, and investigate how the geometry of this semigroup controls continuous-time embeddings. We show that the identity is a non-isolated point of the centralizer precisely when the centralizer contains a non-trivial continuous one-parameter semigroup; such a semigroup need not contain $\phi$. For each logarithmic branch $\mu_k=-\mathrm{Log}\lambda+2\pi i k$, we establish a local sector alternative that characterizes the existence of an embedding with spectral value $\mu_k$. We prove that the set of admissible branch indices is either empty, a finite interval of integers, or all of $\mathbb{Z}$, and give examples realizing all three possibilities. Finally, we describe the full holomorphic centralizer, including the constant map with value equal to the Denjoy–Wolff point.
Read the full abstract

Computer Vision

MESAM: Enhancing SAM2 for Few-Shot Segmentation through Semantic-Visual Memory Generation and Refinement

This paper investigates how to leverage the powerful capabilities of the Segment Anything Model 2 (SAM2) to tackle few-shot segmentation (FSS) tasks. One major challenge lies in the substantial appearance differences between the support and query images, which may diminish the effectiveness of the memory-matching process in SAM2. Through systematic analysis, we find that the memory generated based on the categorical semantic information is robust to this appearance divergence and is more compatible with SAM2. Based on these observations, we propose a new Memory-Enhanced SAM2 (MESAM) method, which first introduces a Semantic-Visual Memory Generation (SVMG) module to separately construct a query semantic memory, a query foreground visual memory, and a support adaptive visual memory. Then, in our designed Interactive Semantic Memory Refinement (ISMR) module, we further enrich the obtained semantic memory with fine-grained visual details, and suppress its misatcivated background responses. Extensive experiments on the PASCAL-$5^i$, COCO-$20^i$, and LVIS-$92^i$ datasets demonstrate that MESAM consistently outperforms existing FSS methods, achieving new state-of-the-art segmentation performance. Our source code is available at: \href{https://github.com/tianyu-zou/MESAM}{https://github.com/tianyu-zou/MESAM}.
Read the full abstract

All authors Tianyu Zou, Shengwu Xiong, Rui Wu, Ruilin Yao, Yi Rong

Group Theory

Bounds on Negligible Thresholds in Group Cohomology

Let $F$ be a field, $K/F$ a field extension with $\Gamma_K:=\text{Gal}(K_\text{sep}/K)$, $G$ a finite group, $j:\Gamma_K\rightarrow G$ a continuous group homomorphism, and $M$ an abelian group. There is an induced map of profinite cohomology $j^*:H^n (G;M)\rightarrow H^n (K;M)=:H^n (\Gamma_K;M)$. An element $u\in H^n (G;M)$ is negligible over $F$ if $j^* (u)=0$ for all such $j$ and field extensions $K/F$. The subset of negligible elements forms a subgroup $H^n (G;M)_{\text{neg},F}\subseteq H^n (G;M)$. The negligible threshold $\text{nthres}^G_F (M)$ is the smallest $N\in\mathbb{N}_{\geq 0}\cup\{\infty\}$ such that $H^n (G;M)_{\text{neg},F}=H^n (G;M)$ for all $n\geq N$. In this paper, we first establish some general bounds for negligible thresholds of arbitrary triples $(G,M,F)$ consisting of a finite group $G$, trivial $G$-module $M$, and field $F$. We then compute the negligible thresholds of an arbitrary group when $F$ is a field of characteristic $p$ and $M=\mathbb{Z}/p^k\mathbb{Z}$. Continuing in the characteristic $p$ setting, we compute these thresholds for $G$ a $p$-group and $M=\mathbb{Z}$, and we compute the thresholds for $G=\mathbb{Z}/\ell\mathbb{Z},M=\mathbb{Z}/s\mathbb{Z}$ for integers $s,\ell\in\mathbb{N}_{>1}$ satisfying certain number theoretic conditions. For fields containing sufficiently many roots of unity, we show that the negligible threshold of odd exponent finitely-generated trivial $G$-modules is finite. For $G$-modules of the form $\mathbb{Z}/p\mathbb{Z}$, $p$ an odd prime, we compute an explicit upper bound for the negligible threshold in terms of $p$ and $|G|$. We conclude by providing explicit computations of the negligible thresholds of elementary abelian $p$-groups with mod-$p$ coefficients, negligible thresholds of cyclic groups, and bounds on mod-$p$ negligible thresholds of general finite abelian $p$-groups.
Read the full abstract

Statistical Methodology

Supervised Factor-Augmented Matrix Regression with Weak Factors

We develop a supervised factor-augmented matrix regression framework for high-dimensional analysis with weak factors. When factor loadings are concentrated on unknown subsets of rows and columns, conventional full-matrix factor extraction may suffer substantial signal dilution, with weakness in one matrix dimension propagating to estimation in the other. We characterize this two-way weak-loading phenomenon for both \(\alpha\)-PCA and projected estimators and show that their loading and factor estimates may deteriorate under weak signals. To address this problem, we propose a sequential response-guided screening procedure that identifies informative rows and columns prior to factor extraction. We establish exact support recovery and show that post-screening factor re-estimation converts the weak-loading problem into a pervasive one relative to the retained dimensions, yielding sharper convergence rates under substantially weaker dimensional growth conditions and consistent factor-number determination. We further establish asymptotic distributions for both unrestricted and bilinear regression estimators based on the estimated matrix factors. Simulations demonstrate substantial improvements in support recovery and factor estimation under weak signals. We illustrate the gains obtained by the response-guided factor construction with two empirical applications.
Read the full abstract

Quantum Physics

Local distinguishability of five five-partite orthogonal product states

The local distinguishability of five orthogonal product states (OPSs) has been studied in several multipartite settings, but the five-partite case still requires a systematic characterization. In this paper, we consider five five-partite OPSs, where any two states are orthogonal on exactly one subsystem. By analyzing the distribution of ten pairwise orthogonality relations over five subsystems, we classify the corresponding orthogonality structures by colored orthogonality graphs. After eliminating identical graphs under relabeling of states and permutation of subsystems, we obtain 443 different graphs in seven genuinely five-partite categories. We prove that the OPSs corresponding to 434 of these graphs can be perfectly distinguished by local operations and classical communication (LOCC), while the remaining 9 exceptional cases fall into two types and require separate analysis. Our results give a characterization of the local distinguishability of five OPSs in five-partite OPSs.
Read the full abstract

All authors Guang-Bao Xu, Jie Zhan, Yu-Guang Yang, Dong-Huan Jiang

Robotics

RevoSim: Scalable Multimodal Tactile Simulation for Dexterous Manipulation

We present RevoSim, a scalable and high-fidelity multimodal tactile simulator for the BrainCo Revo3 hand that jointly models 285 piezoresistive taxels and five visuotactile fingertips. RevoSim efficiently generates synchronized pressure, depth, RGB, and marker-displacement observations; on a single RTX 5090, its depth-conditioned marker-displacement configuration supports up to 4,096 parallel environments. We train in-hand manipulation and grasping policies entirely in simulation, establishing a foundation for future sim-to-real transfer. Project page: https://braincotech.github.io/RevoSim-web/.
Read the full abstract

All authors Ke Zhang, Lixin Xu, Ziyi Wang, Xiyue Dong, Jie Tan, Lianghu Shao, Xianpeng Wang, Chuanyu Li, Renjing Xu

Statistical Computation

Advection-aligned block Vecchia inference for spatio-temporal Gaussian processes

Vecchia approximations enable scalable likelihood-based inference for large Gaussian process models, but their accuracy depends on the choice of conditioning sets. For spatio-temporal processes with advection, target-centered conditioning can omit informative past observations. Motivated by total column ozone observations from the Geostationary Environment Monitoring Spectrometer (GEMS), we propose a data-driven, advection-aligned block Vecchia approximation that incorporates estimated transport direction into temporal conditioning geometry. Observation locations need not lie on a regular grid. An initial advection vector is estimated from the data and used once to select transport-aligned past-time conditioning blocks. The resulting graph is held fixed during optimization, while covariance and advection parameters remain free to vary. This separates neighbor selection from covariance estimation, avoids repeated parameter-dependent neighbor searches, and provides an interpretable connection between conditioning geometry and estimated transport. Batched graphics processing unit (GPU) operations evaluate the likelihood and its gradients. In a matched comparison of daily GEMS datasets from July 2024 and July 2025, advection-aligned conditioning attained a lower fitted Vecchia negative log-likelihood than target-centered conditioning on 53 of 61 days. In six Matérn simulation scenarios with 120 matched eight-hour fields each, advection-aligned conditioning reduced advection-vector root mean squared error relative to target-centered conditioning at the same conditioning-block budget, with the largest reduction under strong transport.
Read the full abstract

Machine Learning

Surrogate Quality Score: A Novel Composite Performance Metric for Model Selection

The growing reliance on high-fidelity, state-of-the-art machine learning based models for engineering decision-making necessitates robust and systematic approaches for model selection. While high predictive accuracy at untrained points is a necessary trait in the final model, other attributes such as model complexity, forecastability, explainability, scalability, and physics-awareness may be equally important or relevant criteria depending on application context and user requirements. In this work, we propose a novel composite performance metric, Surrogate Quality Score (SQS) for model selection based on predictive accuracy and model complexity simultaneously. SQS heuristically integrates these two attributes through a weighted product of normalized measures, yielding a unified score to reflect overall model quality. A user-specified parameter in SQS enables explicit control over the accuracy-complexity trade-off, thereby incorporating human judgement and preference in the model selection process. We extensively discuss and justify the formulation of SQS and demonstrate its utility through an illustrative example. By combining predictive accuracy and model complexity, SQS offers a systematic and objective approach for model selection based on both attributes.
Read the full abstract

Machine Learning

Ordered Policy Optimization

Proximal Policy Optimization (PPO) stabilizes policy updates by clipping the importance ratio between the new policy and the data-collecting policy. The standard rule uses one fixed clipping range for all samples. While this choice offers simplicity and effectiveness, it fails to distinguish statistically harmful extreme ratio values from large, yet tolerable changes. We study how to adapt the clipping radius to the reliability of the induced importance ratio distribution. Our case study demonstrates that reliable updates require controlling these extreme ratios, rather than merely verifying that individual ratios fall within a fixed interval. Motivated by this observation, we propose Ordered Policy Optimization (OPO). In each minibatch, OPO sorts the importance ratios, selects the extreme ratios that potentially have the largest effect on the update, and assigns them adaptive sample-wise radii instead of using the fixed clipping radius. These radii are derived from a prescribed tail profile, which specifies desired target values for the selected extreme ratios. The resulting objective remains first-order and requires only minibatch sorting and a modified policy loss. Experiments conducted on MuJoCo and Atari benchmarks demonstrate that this simple modification achieves strong performance relative to established baselines.
Read the full abstract

Information Retrieval

SkinMind: Equipping Agentic RAG with Human-Inspired Graph Probing for Reliable Dermatology AI

Reliable dermatology AI requires access to verifiable evidence distributed across heterogeneous textual and visual sources. However, conventional retrieval-augmented generation (RAG) typically relies on one-shot retrieval from flat corpora, limiting its ability to adapt retrieval granularity, explore multi-hop relations, and integrate multimodal clinical evidence. We introduce SkinMind, a multimodal agentic RAG framework that equips policy models with human-inspired graph probing for autonomous evidence seeking. SkinMind organizes retrieval as an iterative interaction between an orchestrator and a graph environment. Through eight hierarchical graph-probing tools, the orchestrator progressively identifies relevant entities, inspects node-level information, traverses relations and paths, and retrieves source-text and visual evidence. This progressive-disclosure design allows the model to adaptively determine the retrieval modality, exploration direction, and evidence granularity while constructing a query-specific evidence subgraph. To support autonomous graph probing, we construct SkinMind-KG from a 62M-token corpus of dermatology guidelines and textbooks and 423K images, guided by a dermatologist-informed schema and a conflict-resolution mechanism. The resulting graph contains over 1.0M nodes and 2.0M edges, unifying textual, relational, and visual evidence for retrieval. Across six dermatology VQA benchmarks and two policy-model settings, SkinMind improves the average score over direct answering by 8.70 and 8.89 points with GPT-5.4-mini and Gemini-2.5-Flash, respectively, while consistently outperforming competing RAG methods. Together, SkinMind and SkinMind-KG provide an effective foundation for evidence-grounded multimodal reasoning in dermatology.
Read the full abstract

All authors Jing Hao, Weinan Zhou, Zhihong Lu, Yaling Shen, Yiming Xu, Yefeng Liu, Siyuan Yan, Shuaibo Li, Ming Dai, Junjun He, Zuyi Zhu, Jinjie Gu, Lijun Wu, Jian Wang

High Energy Physics (Theory)

AdS${}_3$/BCFT${}_2$ correspondence and 't Hooft anomaly of U(1) symmetry

We investigate how a U(1) 't Hooft anomaly obstructs symmetry-preserving boundary conditions from the perspective of the AdS${}_3$/BCFT${}_2$ correspondence. By considering a U(1) Maxwell–Chern–Simons system in the bulk spacetime, we show that the BCFT current is not conserved at the boundary. We also derive the general solution for the gauge field in AdS${}_2$ slicing.
Read the full abstract

Software Engineering

DRAFT-Bench: Evaluating Frontier Agents Across the Professional Design Lifecycle

Can frontier general-purpose agents understand client intent, explore creative directions, execute revisions, and prepare production deliverables? We introduce DRAFT-Bench (Design Reliability Across Feedback and Tasks), a lifecycle-oriented evaluation of requirement elicitation, creative exploration, constrained editing, iterative revision, and production handoff. Its Requirement-to-Artifact framework links evolving requirements, protected properties, artifact versions, and delivery obligations. We compare native Codex/Astra and Claude Code across 35 design briefs under 43 conditions, alongside a separate 40-scenario RFQ track. Astra completes 43/43 design conditions and Claude 35/43. All scored core runs receive full structural credit, yet on the 28 core conditions scored for both systems, semantic checks flag violations in 11 for Astra and 5 for Claude; these are model judgments, not expert-confirmed failure rates. RFQ coverage is high, but success ordering changes across cohorts. SKU expansion, revision dependencies, and White ink files expose ambiguity among unmet obligations, missing inputs, and evaluator interpretations. These findings separate execution, requirement compliance, and professional acceptance. Unpaired tracks and different tools preclude claims of RFQ-to-delivery success, factory acceptance, or a pure model ranking.
Read the full abstract

Multiagent Systems

ODD+LLM: Specifying LLM-Agent Behavior for Reproducible Social Simulation

The ODD protocol is the de facto standard for describing agent-based models, but it lacks vocabulary for the components that determine LLM-agent behavior: prompt assembly, memory serialization, parsing, and retry logic. We propose ODD+LLM, a backward-compatible extension adding five documentation blocks grounded in the CoALA cognitive-architecture framework. To show that the gap matters, we conduct a multiverse analysis [Steegen et al., 2016] of a published LLM naming-game simulation, varying six specification choices a reimplementer must make. Variants that are defensible readings of the paper stay within statistical noise of our baseline reconstruction. Missing one instruction that the paper states but is easy to overlook (that the model answers as an outside observer rather than as a player) raises the share of runs settling on the convention agents individually favor from 0.35 to 1.00 and makes convergence three times faster, in a reimplementation that uses a different model from the original. We complement the demonstration with a reconstruction-sufficiency review of 28 published LLM-agent simulations, finding that the details of how prompts are built and how model outputs are parsed are undetermined in 77% of cases, and that no paper fully specifies its main LLM call. ODD+LLM gives the field a standard home for information it already produces but currently scatters across code repositories, supplementary files, and undocumented defaults.
Read the full abstract

Software Engineering

LLM-Augmented Flow for Project Conformation and Task Complexity Evaluation Based on the TLM Framework

The software development process, from inception to a working product or solution, can be divided into two major phases: discovery (Discovery) and delivery (Delivery). The Discovery is further subdivided into High-Level Discovery (HLD) and Low-Level Discovery (LLD). During HLD, management identifies the business solution or product to be integrated into the company's infrastructure. LLD, in turn, produces requirement documentation artifacts, including the Product Requirements Document (PRD), Business Requirements Document (BRD), and Functional Specifications Document (FSD). Story Points (SP) are a widely used modern approach to measuring the effort required to accomplish predetermined tasks, and the artifacts listed above help derive SP more precisely. With the rapid growth of LLM technologies, one of the main questions that arises is whether the whole Discovery phase, or most of it, can be completed with the help of LLM tools. In this paper, we test a framework that structures requirement documentation generation and task complexity scoring, based on the model introduced by (Tran, Lévesque, and Meunier 2004) and executed with the help of agentic tools. The approach is applied to a real-world commercial project in the fintech domain as a feasibility study rather than a validation: development effort was not recorded for the tasks examined, and the resulting complexity scores are therefore reported as the output of the instrument rather than as validated estimates of effort.
Read the full abstract

All authors Maksim Sysoev, Alexander Semenov, Ilia Cheshev, Alexey Gevorkyan, Danil Somov

Robotics

Wellness Robots and the Path to Full Autonomy

This paper introduces the autonomous wellness robot as a distinct category of socially assistive system, defined by its capacity to continuously assess and respond across the seven ICAA dimensions of wellness (see Figure 2). We propose the Care Robot Autonomy Scale (CRAS): a six-level taxonomy (L0-L5) modeled on the automotive automation standard SAE J3016 and evaluated across four wellness-aligned dimensions. We present Ryan, developed initially at the University of Denver and commercialized by DreamFace Technologies, as a case study. NIH/NIA-funded longitudinal evidence (N = 18; 887 hours) shows statistically significant cognitive improvement (SLUMS +2.61, p = 0.015) and a clinically meaningful directional reduction in depressive symptoms (PHQ-9 -2.50, p = 0.073), with no novelty effect. We position Ryan today at CRAS Level 1-2, with emerging Level 3 capabilities in some dimensions, and outline a three-phase roadmap to L5 (full autonomy) by the early 2030s.
Read the full abstract

All authors Mohammad H. Mahoor, Matthew Hessler, Lombe Chileshe, Ola Alsaadi, Eshrat Emamian, Kerstin S. Haring, Richard M. Voyles

Systems and Control

Grid-Oriented Probabilistic Load Forecasting for AI Data Centers

Probabilistic load forecasts for AI data centers need to represent both uncertainty under changing workloads and temporal dependence to support grid operations. This paper develops an open, grid-oriented benchmark using measured GPU power from a 448-GPU AI cluster. Forecasts are evaluated through four stages: probabilistic prediction, online uncertainty calibration, temporal scenario generation, and horizon-wide reserve sizing. Following a change in cluster utilization, a quantile LightGBM forecast achieves only 63.4% coverage for a nominal 90% interval, while online regime-conditioned calibration increases coverage to 87.3%. Even with well-calibrated marginal forecasts, independent sampling produces unrealistic ramp distributions and excessive reserve procurement. Reconstructing temporal dependence from historical forecast errors reduces the discrepancy in ramp statistics and lowers realized reserve cost by 7.4-10.6% across three forecasting tasks. These findings show that marginal forecast quality alone is insufficient for operational decisions and motivate joint evaluation of calibration, temporal structure, and decision outcomes.
Read the full abstract

All authors Qingyi Yao, Ding Lin, Tianqiao Zhao, Meng Yue, Jianhui Wang

Computer Vision

UAV-Based Computer Vision for Cattle Detection and Feed Bunk Spatial Analysis

Beef cattle production is central to Brazilian agribusiness and faces growing demands for efficiency, sustainability, and animal welfare. Precision Livestock Farming (PLF) offers tools to address these challenges, particularly through monitoring feeding-related spatial patterns in confined systems. Feeding management plays a decisive role in animal performance, yet its evaluation still relies largely on subjective, labor-intensive visual inspections that are hard to scale and prone to bias. This work proposes a computer vision framework based on Unmanned Aerial Vehicles (UAVs) to quantify cattle alignment and spatial dispersion at the feed bunk under real-world feedlot conditions. A UAV dataset was collected during flight campaigns at a commercial feedlot in Mato Grosso do Sul, Brazil. Images were manually annotated into cattle head and cattle back classes to support object detection and image-level spatial analysis of cattle distribution. You Only Look Once (YOLO) models, namely YOLOv8 and YOLOv11, were evaluated, with YOLOv8 small used for the downstream spatial analysis. The feed bunk region was segmented using the Segment Anything Model (SAM), and geometric metrics based on detected head positions were developed to characterize alignment, spacing regularity, and clustering. The selected detector achieved 84.30% mAP50 and 35.50% mAP50-95, with an inference time of approximately 2.7 ms per image on an NVIDIA RTX 5060 Ti GPU. The resulting metrics distinguished representative regular and irregular spatial distributions by Root Mean Square Deviation - RMSD (1.06 m and 4.26 m), Coefficient of Variation - CV (42.70% and 96.31%), and clustering (0 and 2 clusters), respectively. These findings indicate that the proposed approach can provide indicators for assessing feed bunk spatial patterns in PLF.
Read the full abstract

All authors Pedro Oliveira Torrente, Diego Giaretta de Paulo, Lucas Pascotti Valem, Jianglong Yan, Everton Castelão Tetila, Jayme Garcia Arnal Barbedo, Liang Zhao

Image and Video Processing

Quantifying System-Level Performance Bottlenecks in Hyperspectral Target Detection

Discerning core performance bottlenecks in hyperspectral target detection is highly challenging given the complex interplay between sub-systems of the imaging chain (scene, atmosphere, sensor and processing). Prior sensitivity and limitation studies offer narrow insights applicable to specific system configurations. To address the need for broader generalization, an analysis framework was constructed to quantify limitations from large ensembles of model scenarios. Key components of the framework include the use of canonical scene representations and the log-weighted area under the ROC curve (wAUC), a metric well-suited for rare target detection. By analyzing thousands of scenarios, parametric thresholds were quantified where detection performance degrades, corresponding to "knees" in performance curves. A knee significance score (KSS) is introduced to assess the relative importance of these thresholds. The framework an airborne VNIR sensor implanted in 28,000 scenarios using seven system parameters spanning the scene, atmosphere and sensor. Results suggest that aerosol visibility in the 2-6 km range is the dominant limiter in detection, followed by subpixel target percentages of 15-35%, and background clutter modeled using a t-distribution with 3-5 degrees of freedom.
Read the full abstract

Materials Science

A stochastic propagation method for large-scale phonon calculations

Large-scale phonon calculations for structurally complex materials are limited by the cubic cost of full diagonalization. We extend the stochastic propagation method (sPM) from electronic systems to lattice dynamics for estimating phonon frequencies, constructing vibrational densities of states, and reconstructing selected phonon polarization vectors by propagating random initial vectors under the dynamical matrix. For finite-range interactions, the leading propagation cost scales linearly with atom number at fixed sampling and propagation parameters. Our error analysis shows that, for nondegenerate modes that are spectrally resolved and have nonzero frequency in the small-error regime, extending propagation yields faster asymptotic reduction of frequency-estimation errors than increasing the number of random initial states alone. Comparisons with direct diagonalization validate the spectral estimates and a selected polarization vector. Convergence tests support the predicted scaling of frequency errors for selected resolved modes.
Read the full abstract

Solar and Stellar Astrophysics

Tidal Braking Does Not Dominate Open-Cluster Slow Rotators

Extended main-sequence turn-offs (eMSTOs) observed in open clusters younger than $2~\mathrm{Gyr}$ are commonly attributed to a spread in stellar rotation rates. Tidal synchronization in close binary systems has been proposed as a possible mechanism for producing slowly rotating stars. To test this hypothesis, we measured the radial-velocity (RV) dispersions of slow-rotator samples associated with the eMSTOs of eight open clusters using ESO VLT/GIRAFFE and SALT/RSS spectroscopy. We extended the analysis to 40 open clusters using Gaia DR3 RV measurements. The observed cluster RV dispersions are generally substantially smaller than those predicted for binary populations dominated by close systems with orbital periods for which tidal synchronization is expected to be effective. In particular, all 40 clusters in the Gaia sample have RV dispersions below the prediction of the $P=10$ d binary-population model. These results suggest that close, tidally synchronized binaries are unlikely to be the dominant origin of slow-rotator populations in young and intermediate-age open clusters, although a contribution from undetected or wider binary systems cannot be excluded.
Read the full abstract

All authors Zixun Zhan, Chenyu He, Weijia Sun, Chengyuan Li

Machine Learning

ResQAdam: Residual-Preserving Adaptive Optimization for Memory-Efficient Quantization Aware Training

Quantization-aware training (QAT) is the most reliable way to obtain accurate low-bit large language models (LLMs), but training with AdamW keeps two dense FP32 states for every weight matrix. We find that low-rank memory-efficient optimizers such as GaLore and LDAdam degrade substantially in QAT. A considerable part of the QAT momentum lies outside its top-r singular subspace, and projection-based methods discard this part at every step. We further observe that this residual remains useful when only its sign is kept. Based on these observations, we propose ResQAdam. It keeps a single BF16 momentum, applies Adam only to the r x r coordinates of the principal subspace, and adds a normalized sign update computed from the residual. We evaluate ResQAdam with two from-scratch QAT methods and three QAT fine-tuning pipelines, with weights quantized down to 2 bits. It reduces optimizer-state memory by 71% to 74% relative to AdamW, stays close to AdamW in most settings, and lowers training throughput by less than 2%.
Read the full abstract

All authors Liqun Ma, Shengkun Tang, Liangyu Wang, Lin Zhao, Wei Luo, Xinjun Niu, Spandan Tiwari, Ashish Sirasao, Kun Zhang, Zhiqiang Shen

Computer Vision

FANO: Anchored Forward-Noise Optimization for Diffusion Inverse Problems

Pre-trained diffusion models have become a widely adopted zero-shot prior for image inverse problems, among which annealed decoupling methods achieve leading reconstruction performance on nonlinear tasks. We find, however, that annealed methods that directly optimize the denoiser input $x_t$ exhibit radial drift in later layers: within-layer optimization drives the noise-component norm of $x_t$ away from its typical value, pushing the denoiser out of its high-performance regime and degrading reconstruction quality. To address this, we propose FANO (Anchored Forward-Noise Optimization). At each annealing layer, FANO fixes the previous denoiser estimate as an anchor and optimizes only the forward noise $e$, whose distribution is known, instead of $x_t$, thereby preserving the forward-diffusion structure of the denoiser input. The standard Gaussian prior on $e$ concentrates its norm near the typical value, and deviations degrade denoiser performance. We accordingly decompose $e$ in polar coordinates into a direction and a radial magnitude, and apply a soft potential to regulate the radius, keeping $x_t$ in the high-density region of the forward distribution. On multiple linear and nonlinear inverse problems, FANO matches the strongest current baselines on linear tasks and achieves consistent improvements on nonlinear tasks.
Read the full abstract

Artificial Intelligence

Glyph: An Interpretability Engine for AI-Driven Scientific Discovery from Clinical Data

Automated scientific discovery has the potential to advance understanding of disease mechanisms and accelerate identification of early disease markers. Manual data annotation is infeasible at population scale, motivating methods that automatically extract pre-defined phenotypes from multimodal and unstructured clinical data. Such pre-defined feature sets determine and limit which associations these approaches can discover. This limitation extends to agentic AI systems that use these tools to query population-scale data. In contrast, foundation models (FMs) trained on large volumes of clinical data can perform tasks beyond their training objectives. This suggests their latent embeddings may represent novel or unexplored features that could yield insights into disease mechanisms, yet such latent features remain difficult to interpret. To address this challenge, we introduce Glyph, an automated multistage framework that extracts features from multimodal and unstructured clinical data using FMs and sparse autoencoders, interprets them as coded clinical concepts, and evaluates their value for clinical scientific discovery. We applied Glyph to abdominal CT volumes, CT radiology reports and laboratory measurements from 41,628 patients across different institutions. We identified 397 predictive concepts supported by 4,463 Bonferroni-significant associations. Moreover, a majority of discovered associations were automatically reproducible in held-out cohorts across different sites. Glyph concepts provided complementary associations to literature-derived markers, leading to gains in predictive capability. More broadly, these findings suggest that foundation model features revealed through interpretability methods could expand clinical understanding of disease. We release Glyph at https://github.com/StanfordMIMI/Glyph.
Read the full abstract

All authors Robbie Holland, Ashwin Kumar, Mateus A. Esmeraldo, David Le, Andrea Fisher, Jiazhen Pan, Louisa Fay, Magdalini Paschali, Vasiliki Sideri-Lampretsa, Ates Fettahoglu, Stefania Moroianu, Philipp Wesp, Ivan Lopez, Tom Nuno Wolf, McKenzie S. White, Eduardo Pontes Reis, Fatima Rodriguez, Curtis P. Langlotz, Vasilis Syrgkanis, Imon Banerjee, Akshay S. Chaudhari, Sergios Gatidis

Artificial Intelligence

Frame: Form-Restricted Authoring with Model-Based Evaluation

Many professional reports reuse the structure of prior editions while updating the content. We formalize this setting as form-restricted report generation: given a sample report and new grounding documents, generate a report that preserves the sample's form while accurately reflecting the new evidence. We introduce FRAME, a benchmark of 120 EDGAR- and arXiv-derived instances across Financial Quarterly, Financial Annual, Academic Simple, and Academic Complex categories, built with a temporal three-snapshot pipeline. We evaluate outputs with three human-validated metrics: structural similarity via semantic tree-edit distance, coverage via information unit recall, and faithfulness via claim-level grounding verification. Across five LLMs, open-source models nearly match proprietary models on structure but lag on coverage and faithfulness; all models perform better on financial than academic reports, and performance declines with longer outputs and more complex grounding.
Read the full abstract

All authors Nan Jiang, Mohammed Mohsin Hussain, Onkar Tiwari, Ankit Jain, Thong Q. Nguyen, Abhishek Agarwal, Vishal Chowdhary

Robotics

HOC2R: Hand-Object Contact Trajectories for Human-to-Robot Motion Transfer

Human videos contain abundant object-manipulation demonstrations, but they do not directly provide motion cues that a robot can follow. We present HOC2R, a Hand-Object Contact trajectory extractor for Robot motion that combines frozen DINOv3 and SAM 2.1 features with learned heads for hand-object association and contact localization. HOC2R supports two modes: count-aware automation that detects, counts, and localizes visible contacts in untrimmed videos, and fixed-two localization for two-finger grippers when contact frames are known. On a 30-task point-annotated benchmark, count-aware automation improves contact-count accuracy from 58.1% to 71.4% and reduces mean localization distance from 19.81 to 10.46 pixels relative to an adapted VRB2 baseline. On seven unseen RGB-D rollouts, a five-task fixed-two checkpoint reduces mean localization distance from 67.67 to 27.02 pixels. Across six robot evaluation sequences, the 30-task fixed-two checkpoint produces four feasible trajectories and succeeds on three tasks, compared with two feasible trajectories and one successful task for each adapted baseline. HOC2R is competitive in the evaluated single-hand scenes, robust to the hand-color shift, and strongest in the evaluated multi-hand scenes. These results support HOC2R as a practical pseudo-labeling method for extracting robot-usable contact trajectories from human demonstrations.
Read the full abstract

Quantum Physics

Optimal Separations Between Randomized and Quantum Query Complexity under Permutation Symmetry

A function $f:\cD\subseteq A^N\to\{0,1\}$ is permutation-invariant if every permutation of its input positions preserves both its domain and its value. We study how this symmetry limits quantum speedups. For $1/6\leq\varepsilon<1/2$, we prove that every such function over an arbitrary finite alphabet $A$ satisfies $R_\varepsilon(f)=O(Q_\varepsilon(f)^2)$, where $R_\varepsilon$ and $Q_\varepsilon$ denote randomized and quantum query complexities with error at most $\varepsilon$. This improves the previous cubic bound (ITCS 2019). We further establish bounds for graph and hypergraph properties, for which vertex relabeling preserves both the promise and the answer. For the same range of $\varepsilon$, we obtain $R_\varepsilon(f)=O(Q_\varepsilon(f)^4)$ for graph properties in the adjacency-matrix model and $R_\varepsilon(f)=O_k(Q_\varepsilon(f)^{2k})$ for $k$-uniform hypergraph properties with fixed $k\geq2$, where the implicit constant depends only on $k$. These improve the previous bounds $R_{1/3}(f)=O(Q_{1/3}(f)^6)$ and $R_{1/3}(f)=O_k(Q_{1/3}(f)^{3k})$, respectively (SICOMP 2024) These bounds are tight for permutation-invariant functions and nearly tight for graph and hypergraph properties. Unstructured search gives a matching quadratic separation with $R_{1/3}(f)=\Theta\!\left(Q_{1/3}(f)^2\right)$. For every fixed $k\geq 2$, we construct explicit families of invariant promise problems on $k$-uniform hypergraphs satisfying $R_{1/3}(f)\geq Q_{1/3}(f)^{2k-o(1)}$.
Read the full abstract

All authors Yunqi Huang, Lvzhou Li, Zekun Ye, Shenggen Zheng

Machine Learning

Composable Causality: A Toolkit for Systematic Time-Series Causal Discovery and Treatment-Effect Benchmarking

Real-world multivariate time series rarely come with verified ground-truth causal structure, so research on time-series causal discovery and treatment-effect estimation depends on synthetic data. Existing tools are either fixed datasets that cannot be modified after release or script-based generators in which each script fixes one combination of mechanism, confounding, and missingness. Neither supports controlled experiments that vary a single property, such as sparsity, noise distribution, or missingness mechanism, while holding the others fixed. Counterfactual evaluation has the same gap, since existing simulators are built per paper with hardcoded mechanisms or produce observational data only. We introduce DTCD (Datasets and Toolkit for Causal Discovery), a data generation library that produces multivariate time series with ground-truth contemporaneous and lagged adjacency matrices. Topology, mechanism, noise, observational mask, and intervention policy are independent components that compose through configuration alone. Paired counterfactual twins, produced by caching the stochastic realization during the factual rollout and re-injecting it during the interventional rollout, recover individual treatment effects to numerical precision. A reference benchmark of eight discovery algorithms on eighty-one configurations and four treatment-effect estimators on eighteen configurations exposes failure modes that aggregate metrics hide. These include the inability of standard discovery methods to handle missing values, regardless of whether the missingness mechanism is informative, and a marginal-variance shortcut (varsortability) that we measure and remove by standardization. Code and data: https://github.com/hferdous/Composable_Causality
Read the full abstract

Robotics

PRISM: Primitive Routing via In-context Skill Mixing for Retentive VLA Adaptation

Vision-Language-Action (VLA) models promise a single language-conditioned policy adaptable to many manipulation behaviors, yet deployment additionally requires skill retention and long-horizon composition: a policy must retain previously acquired skills and reuse them inside composite tasks whose behavior changes across phases. The standard recipe absorbs every new task into one shared update, which provides neither a parameter-level locus for reusable skills nor a mechanism for invoking each skill along the trajectory. We observe that this missing structure is latent rather than absent: per-skill low-rank updates of a pretrained VLA occupy a robustly distinguishable, low-dimensional subspace, and a single composite-task LoRA concentrates roughly half of its Frobenius energy in this subspace, against a near-zero matched random-rotation control. Building on this geometry, we introduce PRISM, which represents reusable skills as parameter-level primitives: each skill family is stored as a dedicated low-rank action-expert adapter in a primitive bank, primitives are augmented with skill-specific low-rank residuals and composed at inference by a per-chunk VLM-conditioned router that reads the prefix representation to track task progress, and the router is supervised by per-chunk labels from a deterministic, task-conditioned, action-derived oracle. During composite training, the vision encoder, the language LoRA, the skill residuals, the router, and the shared action projections and time MLP are updated; every primitive is held frozen, so the base per-skill action-expert adapters are preserved by construction. On RoboCasa365, PRISM achieves 56.43% atomic-skill success and reaches 20.00% on long-horizon composite tasks, while a standard $\pi_{0.5}$-LoRA two-stage fine-tune reaches 17.86% atomic and 17.50% composite success. We additionally validate PRISM on an AgileX PIPER real-robot setup.
Read the full abstract

All authors Jiachen Tao, Gengyu Zhang, Haoxuan Wang, Bing Liu, Yan Yan

Computation and Language

On the Incompleteness of do-calculus for Universal Identifiability

We study the problem of causal effect identification from an arbitrary set of observational and/or interventional distributions, which we call Universal Identifiability (uID). The uID problem generalizes existing classical identifiability (ID) and general identifiability (gID) problems, which considered the identification problem from either mere observational distributions or a specific type of interventional distributions. Prior research has established graphical criteria to determine when a causal effect is identifiable in ID and gID problems. Moreover, algorithms that leverage do-calculus and basic probability manipulations have been proposed to compute these effects. Consequently, it has been shown that do-calculus is both sound and complete in these problems. Despite these well-established results, a formal definition of what it means for the do-calculus to be complete does not exist. We propose a rigorous formalization for the completeness of the do-calculus and show that the completeness results in previous work remain valid. However, for more general problems, such as uID, the rules of do-calculus are no longer complete. This is perhaps the reason why the uID problem has remained open thus far.
Read the full abstract