跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+197新增论文
+151其中 AI 相关
+16已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-21 12:19 · 全站最近入库:2026-11-21 12:19各信源进度 →

统计口径与数据新鲜度

统计 2026-11-21 06:21 — 2026-11-22 06:21(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-21 12:19最近发现新增版本:2026-11-21 12:19库内论文最新更新时间:2026-11-20 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

29,158 已收录论文24,181 AI 相关8,992 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 21,490 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-21 12:17(北京时间)
新论文 · v1arXiv:2611.20812

The Cybersecurity Dataset Observatory: Which Data for AI in Cybersecurity?

Francesco Cerasuolo, Gabriele Mangiacapre, Antonio Pescapè

原文摘要

AI is increasingly used for cybersecurity tasks such as intrusion and malware detection, in enterprise, industrial, vehicular and mobile networks. Its effectiveness depends on data: models need representative network traffic, both to be trained and to be evaluated realistically. Public datasets of cybersecurity network traffic, however, are scattered across laboratory pages, repositories and aggregators, and are continuously released, corrected and withdrawn: 92 of the 234 datasets we catalog were released in 2023 or later. At this pace, surveys that list them are outdated soon after publication, since they describe the field only as it was when they were written. We present the Cybersecurity dataset observatory, a public and continuously updated catalog of cybersecurity datasets based on network traffic. New datasets are added as they are found or submitted, every record is periodically re-checked against its original source, to which it links rather than hosting the data, and every change is recorded in a public change log. This paper documents the first release: we define a traffic-centered scope and a seven-axis taxonomy, and review 234 datasets (204 original captures and 30 derivatives) across 14 domains. The review shows a field that grows quickly but unevenly, where 91 datasets distribute only processed records, a few laboratories produce a large share of the data, and UAV, wireless and V2X networks have little public evidence, while 5G has few real captures. We then relate the catalog to common uses in AI research, outline a procedure for selecting datasets, and discuss benchmark pitfalls and open gaps.

cs.CR · cs.AI

新论文 · v1arXiv:2611.20810

Prediction-Driven Contract Matching with Controlled Overbooking for Cloud-Edge DDoS Scrubbing Markets

Yue Luo, Houyi Qi, Minghui Liwang, Xinlei Yi, Han Jiang, Xuejiao Zheng

原文摘要

Cloud-edge distributed denial-of-service (DDoS) scrubbing requires ex-ante capacity commitments under uncertain traffic, heterogeneous budgets, resource costs, and fulfillment risks. We develop a prediction-driven two-stage contracting mechanism coordinating business entities (BEs), edge scrubbing nodes (ESs), and a cloud scrubbing center (CS) through BE-ES (B&E) main-service and ES-CS (E&C) supplementary contracts. In the first offline phase, Transformer forecasts generate candidate reserved scrubbing rates, generalized Nash bargaining determines candidate prices, and interim many-to-one B&E matching permits controlled overbooking under scenario-based aggregate ES shortfall-risk assessment. In the second offline phase, formal E&C reservations are established subject to capacity and payment constraints; their finalized rates and prices update ES-side proxy scores and guide localized risk-aware B&E contract adjustment. During online fulfillment, an edge-first, cloud-assisted weighted proportional allocation policy allocates scrubbing capacity under finalized contracts without rematching, renegotiation, or additional cloud reservations. Experiments with CICDDoS2019-derived workloads show that the proposed mechanism satisfies the prescribed empirical shortfall-risk criterion across all 100 evaluated horizon-instance pairs, whereas the NoRisk ablation violates it in all 100. Runtime measurements indicate lightweight online fulfillment relative to offline contract formation. The evaluated configurations also exhibit limited overall scrubbing coverage and near-zero realized cloud utilization.

cs.NI

版本更新 · v2arXiv:2608.00005

RubricReviewer: From Direct Critique to Rubric-Driven Review via Agentic Evidence Synthesis and Human-Aligned Generation

Shuyu Guo, Wenxiang Hu, Yuyue Zhao, Lun Zhou, Yougang Lyu, Xiaohui Yan, Zhaochun Ren

原文摘要

Peer review at major venues is under unprecedented submission pressure, motivating the use of large language models (LLMs) as review assistants. Existing LLM-based reviewers, however, face two structural limitations. First, they map manuscripts directly to reviews, leaving the underlying rubric implicit and entangling its derivation with the judgement. Second, the prevailing paradigms each capture only half of a good review: training-free agents gather broad evidence but produce undirected critiques, while training-based reviewers inherit human discriminative judgement together with its noise and uneven coverage. We introduce RubricReviewer, a fully rubric-driven framework that addresses both limitations. It makes rubric generation an explicit intermediate step, so that both review generation and the final assessment are conditioned on paper-adaptive rubrics. It further combines a training-free agent (Scout) that gathers external evidence with a human-aligned trained model (Aligner) that consumes this evidence, fusing the strengths of both supervision sources. Experiments on real-world submissions show that RubricReviewer produces reviews that are markedly more comprehensive and more discriminative than prior systems, and exhibits the strongest robustness against adversarial prompt-injection attacks. Ablation studies further confirm the necessity of each component.

cs.CL · cs.AI

版本更新 · v2arXiv:2609.16914

Regular Expression Matching with Ordered Linear Backreferences

Philip Bille, Inge Li Gørtz, Rikke Schjeldrup Jessen

原文摘要

A regular expression with backreferences (rewb) specifies a set of strings formed by characters combined with concatenation, union, star operators, and backreferences. A backreference consists of a capturing group $(\cdot)_i$ and one or more references $\backslash i$. The substring matched by the reference must match the same substring that was matched by the corresponding capturing group. We focus on a subclass, which we call ordered linear rewbs where the references and capturing groups occur only outside star and union operators. Furthermore, we require that all references to a capturing group occur before the next capturing group in the rewb. In full generality, rewb matching is NP-complete, but efficient solutions exist for various subclasses. For linear rewbs with one capturing group and $k$ references, Uezato~[CPM 2026] gave an $O(k n^2 m^2)$ time and $O(n^2m^2)$ space algorithm, where $m$ is the length of the regular expression $R$ and $n$ is the length of the string $Q$. For the special case of $k=1$, Nogami and Terauchi~[MFCS 2025] gave an $O(n^2m^2)$ time and $O(n+ m^2)$ space algorithm. On the other hand, Nogami, Nakamura, and Terauchi~[arXiv 2026] gave a conditional lower bound, showing that we cannot solve the problem in $O(n^{2-ε} \mathrm{poly}(m))$ for any $ε> 0$ assuming the orthogonal vector hypothesis. Our main result is a new improved algorithm that runs in $O(n^2m+n^2k\log n)$ time and uses $O(nm)$ space, and handles the more general case of ordered linear rewbs with multiple capturing groups.

cs.DS

新论文 · v1arXiv:2611.20805

An Efficient Training Paradigm for Image Restoration via Improved Fréchet Loss

Shanyuan Jie, Wu Shi, Xinqi Lin, Kainan Yan, Zian Wang, Yingxue Pang, Yan Wang, Shijie Zhao 等 9 位作者

原文摘要

Diffusion-based image restoration achieves strong perceptual quality but typically relies on iterative sampling, leading to slow inference. Existing single-step methods often use diffusion distillation or adversarial training, complicating optimization. To simplify training, we propose FDIR, a framework for training single-pass generative restoration models through regression-style supervised optimization using directly computable objectives. To realize this paradigm, FDIR employs the Fréchet distance to match the feature distributions of restored and real images in frozen representation spaces, providing explicit distribution-level supervision. To better accommodate image restoration, we incorporate local patch features to capture fine-grained restoration discrepancies. Moreover, we introduce instance-adaptive statistics by combining the feature moments of each SR-HQ pair with synchronized historical statistics, allowing distribution matching to incorporate input-specific correspondence while retaining distribution-level context. FDIR optimizes only the LoRA parameters of the restoration backbone. Experiments demonstrate strong perceptual restoration quality and efficient training, suggesting a simpler paradigm for generative image restoration.

cs.CV

版本更新 · v2arXiv:2408.05033

Approximate Distributed Monitoring under Partial Synchrony: Balancing Speed and Accuracy

Borzoo Bonakdarpour, Anik Momtaz, Dejan Ničković, N. Ege Saraç

原文摘要

In distributed systems with processes that do not share a global clock, \emph{partial synchrony} can be achieved by clock synchronization that guarantees bounded clock skew among all processes. Existing solutions for distributed runtime verification under partial synchrony against temporal logic specifications are exact but suffer from significant computational overhead. In this paper, we propose an \emph{approximate} distributed monitoring algorithm for Signal Temporal Logic (STL) that mitigates this issue by abstracting away potential interleaving behaviors. This conservative abstraction enables a significant speedup, albeit with a tradeoff in precision. We address this tradeoff by combining our approximate monitor with its exact counterpart, gaining efficiency without sacrificing precision. Our framework supports \emph{offline} monitoring of both future and past STL, as well as \emph{online} monitoring of past STL. We evaluate our approach with multiple experiments, showcasing its efficacy in both real-world applications and synthetic examples.

cs.LO

新论文 · v1arXiv:2611.20803

What Do Digital Traces Predict? Task Effects, Student Differences, and Critical Online Reasoning

Maxim Konca, Alexander Mehler

原文摘要

Digital assessment traces are increasingly used to predict student performance, often evaluated by correlating predicted and observed scores pooled across tasks. However, such correlations may reflect between-task differences rather than performance differences among students within tasks. To investigate this, we analyzed up to 527 browsing sessions from 424 university students completing open-Internet Critical Online Reasoning tasks and assessed the cross-task stability of their ratings. We evaluated the same out-of-fold predictions within and across tasks, compared three validation tiers up to leave-one-task-out, and tested incremental predictive contributions of consultation order and relational structure against matched permutation and randomization controls. None of the twelve prespecified primary tests was significant: on held-out tasks, consulted-page features showed no significant association with relative student performance, nor did observed order or relational structure add predictive value. Pooling observations across tasks substantially increased correlations, particularly for factual correctness, whose scores showed limited evidence of stable between-student differences across tasks. However, sentence embeddings yielded modest positive correlations with generic response quality on held-out tasks, without establishing added predictive value from observed order or relational structure. Our findings show that pooled predictive correlations may not reflect within-task student differences, highlighting the need for within-task evaluation and criterion generalizability audits.

cs.CL · cs.CY · cs.LG

版本更新 · v4arXiv:2406.10891

Sfumato Supervision: Learning Instance Segmentation with Noisy Labels

Moshe Kimhi, Ehud Rivlin, Chaim Baskin

原文摘要

Instance segmentation requires precise object masks, yet both human annotators and promptable models can produce unreliable contours. A high prediction loss near a contour does not necessarily indicate an annotation error: valid edges can also be uncertain. Inspired by Leonardo da Vinci's \emph{sfumato}, which softens transitions between a subject and its surroundings, we introduce \textbf{\texttt{CB-LNL} (Copula-Boundary Learning with Noisy Labels)}. Rather than judging each contour pixel alone, CB-LNL combines foreground evidence inside the mask with background evidence outside it to guide local supervision and annotation repair. We ground this approach in an audit of paired COCO and LVIS annotations, and introduce benchmarks covering human annotations, weak-prompt SAM2 masks, and controlled corruptions in natural and driving scenes. Controlled comparisons show improved training on procedural contour noise and fewer harmful edits than direct teacher-mask replacement. Paired-seed and frozen-prediction evaluations reveal how these benefits depend on the annotation source, supporting separate assessment of category and contour reliability.

cs.CV · cs.LG

新论文 · v1arXiv:2611.20797

Correspondence-Free Camera Pose Estimation by Matching Moments

Subhransu S. Bhattacharjee, Dylan Campbell, Rahul Shome

原文摘要

The blind perspective-$n$-point (PnP) problem is to estimate the pose of a calibrated camera from 2D and 3D point sets with unknown correspondences. We show that, for complete, noiseless observations, fourteen permutation-invariant image moments up to order four identify a fixed camera pose for almost every 3D point set containing at least twenty points. Our learning-free method, MomentPnP, matches these moments between the observed 2D points and the projected 3D points to initialize a camera pose without correspondences and refines the pose and correspondence weights using expectation--maximization. Across the test sets of three benchmarks, our method reduces median rotation error by $5$--$170\times$ relative to the best learned baseline, both with and without outliers. We recover accurate poses even with as many outliers as inliers. On dense ETH3D scans, pose accuracy improves as the point sets grow, even as individual correspondences become harder to resolve.

cs.CV · cs.LG · cs.RO · eess.IV · math.OC

版本更新 · v3arXiv:2501.16698

3D-MoE: Towards Spatial Intelligence with Mixture-of-Experts for 3D Reasoning and Action Generation

Yueen Ma, Zenglin Xu, Irwin King

原文摘要

Spatial intelligence, encompassing 3D perception and reasoning, is the essential next frontier of AI. Scaling current 3D vision-language models (VLMs) that rely on dense Transformers for spatial tasks incurs prohibitive computational costs. In this paper, we introduce 3D-MoE, a 3D VLM leveraging an efficient mixture-of-experts architecture with a modality- and spatial-context-aware probabilistic routing scheme, stably cultivated by a novel routing curriculum. To seamlessly extend 3D-MoE to embodied AI, we integrate a diffusion-based action head, Pose-DiT, transforming 3D-MoE into a 3D vision-language-action (VLA) model. By employing a rectified flow framework, Pose-DiT generates precise 6D pose actions in a single sampling step. Extensive experiments demonstrate that 3D-MoE achieves superior performance on diverse 3D vision-language benchmarks with drastically fewer activated parameters and yields higher success rates while enabling real-time inference for robot manipulation tasks. Code is available at https://github.com/yueen-ma/3D-MoE.

cs.CL · cs.CV · cs.RO

新论文 · v1arXiv:2611.20795

Speed-Robust Scheduling for Total Completion Time

Moritz Buchem, Nicole Megow, Morten Weber

原文摘要

We study two-stage speed-robust scheduling for minimizing total completion time. Jobs must be partitioned into $M$ bags before an adversary reveals the speeds of at most $M$ machines. Each bag is then assigned to one available machine, and the jobs on each machine may be processed in any order independent of the bags. We compare with an optimal schedule that knows the speeds in advance and assigns jobs individually. This model has been studied for makespan minimization. We initiate its study for total completion time, as suggested by Stein and Zhong (TALG 2020). For an unknown number of identical machines, we show that the optimal deterministic robustness factor is exactly~$13/12$, both for unit-size jobs and for arbitrary processing times. For infinitesimally small jobs, the factor lies between $31/30\approx1.0333$ and $28/27\approx1.0370$, well below the optimal makespan factor of about $1.2071$. General speeds make the problem strictly harder: no deterministic algorithm is better than $1.2669$-robust, already for infinitesimally small jobs. We give a $25/16$-robust algorithm for arbitrary processing times and improve the guarantee to $1681/1152\approx1.4592$ for infinitesimally small jobs. Our main tools are a layer decomposition that extends guarantees from unit-size jobs to arbitrary processing times when one assignment satisfies them for every prefix in non-increasing processing-time order, and a threshold analysis of the bag counts produced by greedy assignment rules.

cs.DS · cs.DM

版本更新 · v2arXiv:2610.41398

Distilling Knowledge from Search: Decoder Model Factory for Quantum LDPC Codes

Anil Bilgin, Wanda Hou, Pranav Srinivas Murali, Yi-Zhuang You

原文摘要

We introduce GraphMP, a learning-based message-passing decoder for quantum low-density parity-check codes that transfers without fine-tuning across code families, code sizes, physical error rates, and noise models. Its local message passing keeps computation scalable as the decoding graph grows. We further embed GraphMP in a beam-search algorithm that progressively refines candidate error patterns, trading additional decoding time for lower logical error rates. To turn this decoding effort into lasting improvements, we propose the decoder model factory, a paradigm for training successive models on data collected by earlier models during decoding. In our implementation, verified search trajectories provide reusable training data for distilling knowledge into subsequent models. Across generations, the model factory yields a search-enhanced decoder that is broadly competitive with Relay-BP. On the largest code block, which encodes $4612$ logical qubits, it achieves a lower logical error rate and about $20$ times the decoding throughput of the standard Relay-BP baseline. These results motivate a coupled quantum-classical architecture with two interacting cycles: rapid decoding protects quantum information, while slower learning adapts the models to changing noise conditions. The long-term vision is a scalable decoder model factory that continually produces high-performance, low-latency decoders for evolving quantum hardware.

quant-ph · cond-mat.dis-nn · cs.AI

版本更新 · v6arXiv:2603.09632

X-GS: An Extensible Framework for Perceiving and Thinking with 3D Gaussian Splatting

Yueen Ma, Zenglin Xu, Irwin King

原文摘要

3D Gaussian Splatting (3DGS) has emerged as a powerful technique for novel view synthesis, subsequently extending into numerous spatial AI applications. However, most existing 3DGS methods operate in isolation, focusing on specific domains. In this paper, we introduce X-GS, an extensible framework that integrates previously isolated 3DGS methods into the perception module of a VLM for spatial tasks, with two major components: the $\textit{Perceiver}$ and the $\textit{Thinker}$. The $\textit{Perceiver}$ performs online 3DGS-based SLAM with semantic distillation and outputs semantic Gaussians from unposed video streams. It leverages recent vision foundation models for stronger geometric priors, and we introduce three novel optimizations to improve semantic distillation efficiency. The $\textit{Thinker}$ interfaces diverse VLMs with these semantic Gaussians, unlocking spatial multimodal capabilities such as 3D visual grounding and scene captioning. Experimental results on diverse benchmarks demonstrate the efficiency and newly unlocked multimodal capabilities of the X-GS framework. Code is available at https://github.com/yueen-ma/X-GS.

cs.CV · cs.CL

新论文 · v1arXiv:2611.20791

QTOP: Scaling Humanoid Motion Tracking with Quantile Critics

Markus Peschl, Dana Kianfar, Daniel Dijkman

原文摘要

Learning human motion tracking policies through reinforcement learning (RL) has emerged as a core capability that underpins a majority of recent progress in humanoid whole-body control. However, achieving generalization to unseen motions at high fidelity requires jointly scaling training to thousands of GPU and dataset hours. This work hypothesizes that motion tracking is inherently compute-bound because training on a diverse set of motions introduces learning interference, similar to multi-task RL. To tackle this, we propose Quantile Tracking Optimization (QTOP), which improves sample efficiency by introducing a distributional advantage estimate conditioned on state and reference pairs. QTOP outperforms on and off-policy RL baselines both in simulation and real-world tracking scenarios on multiple datasets by up to 10\%. Furthermore, we establish a scaling law for QTOP and find that, aside from scaling predictably, dataset size has minimal impact on performance for moderately sized GPU hour regimes.

cs.RO

新论文 · v1arXiv:2611.20788

VA-Assembly: View-Aligned Score Guidance for Training-Free 3D Part Assembly

Yeon-Jae Kim, Sangwon Kang, Byung-Kwon Min

原文摘要

Image-to-3D foundation models have improved 3D reconstruction, but typically produce monolithic representations without explicit part-level structure. Part-aware generation and learning-based assembly methods often require part-level supervision, category-specific priors, or paired part-layout training data. We instead consider a practical but underexplored setting in which independently prepared or generated 3D parts have no predefined inter-part relationships, shared coordinate frames, or consistent scales, and must be assembled into a reference-consistent object from a single image. We present VA-Assembly, a training-free framework for 3D part assembly through part-level transformation estimation using view-aligned diffusion guidance. Given a reference image and fixed 3D part geometries, VA-Assembly optimizes each part's translation, rotation, and scale. We condition a frozen view-conditioned diffusion model on same-view renderings of a reference-derived monolithic 3D representation, producing score gradients geometrically aligned with the part configuration. This guidance provides multi-view structural cues under large initial misalignment. By repurposing pretrained visual-structural priors as transformation-level supervision, VA-Assembly requires no annotated part layouts, predefined part taxonomies, category-specific assembly training, or bounding-box-based layout initialization. Experiments compare VA-Assembly with geometric registration and compositional layout baselines using each baseline's supported transformation components, showing improved part-wise transformation estimation. Ablation studies show that view-aligned conditioning, rather than diffusion guidance alone, is critical for reliable transformation gradients. Updating only part transformations preserves explicit component separability and supports component-level editing, replacement, and reconfiguration.

cs.CV

新论文 · v1arXiv:2611.20787

PhaseCLIP: Phase-Aware Frequency-Domain Adaptation of Vision-Language Models for Few-Shot Remote-Sensing Classification

Mehmet Sergen Catal, Abdurrahman Gumus, Mehmet Zübeyir Ünlü

原文摘要

Remote-sensing (RS) image scene classification aims to assign a semantic category to an overhead image, and in practice must often do so from only a handful of labelled examples, since annotation requires expert interpretation and the resulting labels transfer poorly across sensors and regions. Recent progress has come from adapting vision-language models (VLMs) such as CLIP, through learned prompts or feature adapters. Although effective, these methods construct their representation entirely in the spatial domain. RS scenes are dominated by periodic texture and regular geometry, which the Fourier spectrum describes compactly through an amplitude component, carrying texture and contrast statistics, and a phase component, carrying spatial structure. Existing frequency-domain adaptation of VLMs either exploits amplitude alone or does not separate amplitude from phase, leaving the structural component unused. In this article, we propose PhaseCLIP, which inserts phase-capable spectral modules and a feature-gating module into a frozen CLIP visual encoder and trains only 1.30% of its parameters. Because the amplitude and phase pathways share an identical parameter layout, a single weight set is learned on phase and transferred exactly to amplitude, giving a two-stage scheme that adds no parameters. Experiments on five public RS datasets, EuroSAT, PatternNet, UC Merced, LSRS and RSSCN7, against eight baselines under a matched protocol, show that PhaseCLIP attains 84.28% mean accuracy and is the best method in 23 of 30 dataset-by-shot cells, with phase contributing 1.93 points overall and 4.36 points on the most difficult dataset. Because the modules act on the spectrum of image patches and transformer tokens and leave the text encoder untouched, the approach can be extended to other RS sensors and tasks, and to other domains in which class evidence lies in spatial structure.

cs.CV

版本更新 · v2arXiv:2611.00921

Solvable dynamical model of in-context learning

Ruoran Bai, Haiping Huang

原文摘要

Pretrained Transformer architectures display a surprising in-context learning ability without additional training. How the learning process guides the network to achieve this intriguing capability remains unsolved. In this study, we focus on a single-layer single-head linear attention model, where an empirical risk minimization is defined for learning. Given a finite number of training examples, we determine the exact solution of network-parameter dynamics, which is linked to the training error and generalization error, characterizing how memorization and generalization emerge from learning and how these transitions are related to curvatures of the empirical risk landscape. Furthermore, we calculate a data-dependent free energy, from which the training and generalization errors in the long time limit of learning dynamics can be well predicted. Our work adds insights toward understanding the mechanism of in-context learning especially from a learning dynamics perspective.

cs.LG · cond-mat.dis-nn

版本更新 · v2arXiv:2605.13534

Scaling Retrieval-Augmented Reasoning with Parallel Search and Explicit Merging

Jiabei Liu, Wenyu Mao, Junfei Tan, Chunxu Shen, Lingling Yi, Jiancan Wu, Xiang Wang

原文摘要

Deep search agents have proven effective in enhancing LLMs by retrieving external knowledge during multi-step reasoning. However, existing methods often generate a single query for retrieval at each reasoning step, limiting information coverage and introducing high noise. This may result in low signal-to-noise ratios (SNR) during search, degrading reasoning accuracy and leading to unnecessary reasoning steps. In this paper, we introduce MultiSearch, an RL-based framework that addresses these limitations through multi-query retrieval and explicit merging of retrieved information. At each reasoning step, MultiSearch generates queries from multiple perspectives and retrieves external information in parallel, expanding the scope of relevant information and mitigating the reliance on any single retrieval result. Then, the agent consolidates and refines retrieved information at the merging process, improving the SNR and ensuring more accurate reasoning. Additionally, we propose a reinforcement learning framework with a multi-process reward design to optimize agents for both multi-query retrieval and information consolidation. Extensive experiments on seven benchmarks demonstrate that MultiSearch outperforms baseline methods, enhancing the SNR of retrieval and improving reasoning performance in question-answering tasks.

cs.AI

新论文 · v1arXiv:2611.20786

ThermoSight: From Appearance to Internal Thermofluid Fields via Physics-Guided Neural Reconstruction

Yuhao Fang, Chenglin Lu, Hefei Ling, Chun Lou

原文摘要

Most neural reconstruction methods focus on reproducing scene appearance, yet inferring physical states from visual observations is essential for machine understanding of the physical world. Reconstructing the internal temperature and soot fields of radiating thermofluids such as flames is challenging yet essential for improving energy efficiency and assessing heat exposure and fire spread. In flames, temperature and soot concentration jointly govern visible emission, and soot self-absorption alters the light that reaches each camera. With limited views, changes in the two fields can offset one another in the recorded RGB images, making temperature and soot difficult to disentangle. ThermoSight, a physics-guided neural framework for scientific imaging, jointly reconstructs continuous three-dimensional temperature and soot volume fraction fields from limited calibrated flame images without field-level ground truth during optimization. A differentiable spectral model links these fields to measured RGB through thermal emission, self-absorption, and calibrated camera responses. Temperature-dependent emission spectra and wavelength-dependent attenuation constrain shared fields across RGB channels, reducing temperature-soot ambiguity. Chromaticity consistency, observation-derived spatial constraints, and soot regularization complement intensity fitting. In a synthetic ablation, wavelength-resolved attenuation reduces temperature RMSE in the hot, sooty region from 234 to 112 K relative to a three-channel approximation, with nearly identical RGB PSNR (65.23 vs. 65.24 dB). Measured data yield coherent fields and competitive RGB reconstructions. These results demonstrate that physics-guided neural fields recover interpretable internal physical quantities beyond visually consistent appearance.

cs.CV · physics.app-ph

新论文 · v1arXiv:2611.20784

On homogeneous hypertournaments

Matěj Konečný

原文摘要

Sullivan recently constructed continuum many countable homogeneous $n$-hypertournaments for every $n\ge3$ by forbidding configurations he calls pair-cancelling. We show that, for every family $\mathcal F$ of pair-cancelling $n$-hypertournaments, the class of all linearly ordered finite $\mathcal F$-free $n$-hypertournaments is Ramsey and has the ordering property. In particular, freely ordered $\mathbf{H}_4$-free $3$-hypertournaments are Ramsey, answering a question of Cherlin, Hubička, Nešetřil, and the author from 2021. For every odd $n\ge3$, we give a simple Ramsey expansion with the expansion property for even $n$-hypertournaments by naming an $(n-1)$-hypertournament and a free linear order. We also show that, for $n\ge6$, no unbounded hereditary strong amalgamation class of finite $n$-hypertournaments has EPPA. Moreover, we prove the failure of EPPA for all Sullivan's classes with $n\ge4$, and for the even $5$-hypertournaments. We conjecture that even hypertournaments and classes omitting a family of pair-cancelling hypertournaments, together with the homogeneous cyclic order in arity three, are the only countably infinite homogeneous $n$-hypertournaments for $n\ge3$. We give some supporting evidence.

math.CO · cs.DM · math.LO

版本更新 · v2arXiv:2608.05615

ALTER: Modeling Longitudinal Changes via Regional Differencing for 3D CT Report Generation

Dongchen Li, Jitao Liang, Wei Li

原文摘要

Computed tomography (CT) is widely used for clinical diagnosis and longitudinal follow-up, yet automatically generating accurate and complete radiology reports from three-dimensional (3D) CT remains challenging. Existing methods improve fine-grained correspondence between images and text by modeling anatomical regions, but remain centered on the current examination. Consequently, patient-specific longitudinal changes within individual regions remain insufficiently modeled. Meanwhile, interval changes are often distributed across multiple anatomical regions, complicating a coherent assessment of the overall longitudinal state. We propose Anatomically Localized Temporal Evidence Representation (ALTER) to address these limitations. Global Prior Integration (GPI) incorporates the prior CT and report to establish historical context for the current examination. Regional Proxy Differencing (RPD) enables each current anatomical region to retrieve a historical proxy from a single shared encoding of the prior volume and to derive localized interval evidence. Interval Change Fusion (ICF) further combines current abnormality states with region-distributed differences, converting their joint representation into change-aware soft prompts that guide report generation. ALTER achieves state-of-the-art results on most evaluation metrics across the RadGenome-ChestCT validation and CTRG-Chest-548K test sets. Code and data preprocessing details are available at https://github.com/peytonkarlie/ALTER/tree/main.

cs.CV

新论文 · v1arXiv:2611.20782

Baryonic Keys and the Inheritance of Crypto-Assets: Revisiting the Tales From the Crypt Protocol

Frédéric Prost

原文摘要

The inheritance of crypto-assets faces a structural obstacle: ownership is the knowledge of a key, so the key either disappears with its owner or has to be shared beforehand. The Tales From the Crypt Protocol (TFCP) addressed one part of the problem, the distributed and privacy-preserving acknowledgment of a death by a blockchain. We revisit it with a new primitive. A \emph{baryonic key} is a physical object from which bits are obtained only through a measurement, are never stored digitally, and which exists in exactly $n$ copies, where $n$ is neither zero nor unbounded. We define the primitive, outline a candidate instantiation based on a replicated micro-fractured relief, and state its limits: digitising a copy is slow and tamper-evident, not impossible. We also outline other uses of the primitive, from post-quantum pre-shared keys to common coins for Byzantine agreement. We then give a threat model for TFCP, review the attack models of the literature that apply to it, and show where it resists and where it does not. We propose TFCP-B, in which decrypting the will requires both a secret shared among registrars and the response of a baryonic key, and in which heirs claim their share through hash commitments, without any prior key or wallet. We detail each change from TFCP to TFCP-B, with its rationale and its price. TFCP-B removes the anonymity loss caused by colluding registrars and blocks several false-alarm attacks. It does not remove the shallow-account attack, remains exposed to transaction censorship during the deliberation period, and introduces new assumptions on custodians and reading devices. The security arguments are informal and the physical figures are design estimates.

cs.CR

版本更新 · v2arXiv:2607.21186

On the effects of emulated quantum layers on CNN performance and explainability for OCT and dementia imaging

Guillermo Rubiños Rodríguez, Martín Ottavianelli, Mateo Alonso, Gonzalo Blázquez Gil, Boris-Stephan Rauchmann, Pablo Díez-Valle, Sergio Altares-López

原文摘要

Hybrid quantum-classical convolutional neural networks have been proposed as an alternative to classical deep learning models. However, executing network components on quantum hardware imposes fundamental limitations, such as gradient estimation through repeated measurements, and increasing the number of qubits can lead to trainability problems such as barren plateaus. Moreover, the reported success of these hybrid networks has been linked to quantum architectures that can be efficiently simulated on classical computers. We investigate whether small, classically simulated quantum circuits can serve as useful components within larger models, as a substitute for purely classical layers. We present a systematic comparison between a Hybrid Quantum-inspired CNN (HQiCNN) and standard CNN architectures, where the models differ only in a single intermediate layer: an emulated quantum circuit in the HQiCNN and classical counterparts in the CNN. The models are evaluated on two medical imaging datasets while varying training set size, learning rate, batch size, circuit entanglement and observable se. Under a strict parameter budget, the HQiCNN outperforms the classical baselines. When the classical layer is given more parameters, overall performance becomes comparable, with the HQiCNN showing improvements only in the intermediate-data regime. Removing entanglement yields comparable performance and richer observable sets yield no performance improvement. Finally, we introduce two SHAP-based metrics for comparing model explanations, $|$SHAP$|$IoU and EMD$_{\mathrm{pos}}$. They show that all architectures attend to anatomically plausible regions, and that the spatial attention agreement between them increases with training set size. These results indicate that, under specific data regimes, quantum-inspired hybrid models are a competitive alternative to classical CNNs for medical image classification.

quant-ph · cs.AI · cs.LG

版本更新 · v6arXiv:2606.18216

Zone of Proximal Policy Optimization: Teacher in Prompts, Not Gradients

Byung-Kwan Lee, Ximing Lu, Shizhe Diao, Minki Kang, Saurav Muralidharan, Karan Sapra, Andrew Tao, Yu-Chiang Frank Wang 等 9 位作者

原文摘要

Knowledge distillation transfers a teacher's competence to a small student but is brittle in the small-student regime: forcing the student to imitate logits from a much larger teacher hurts generalization on benchmark families beyond the training corpus. Reinforcement learning (RL) avoids logit imitation by training on the student's own rollouts. However, questions on which every rollout fails yield zero advantage and are silently discarded, and injecting a stronger teacher's response into the policy gradient to recover them breaks the on-policy assumption and induces drift. We introduce Zone of Proximal Policy Optimization (ZPPO), inspired by Vygotsky's zone of proximal development. ZPPO keeps the teacher inside the prompt rather than the policy gradient. On hard questions, where the student's mean rollout accuracy is below half, ZPPO constructs two reformulated prompts. A Binary Candidate-included Question (BCQ) pairs one correct teacher response with one incorrect student response as anonymized candidates the student uses as references. A Negative Candidate-included Question (NCQ) aggregates the student's wrong rollouts into a single prompt to surface their shared failure modes. A prompt replay buffer keeps each hard question eligible for re-sampling until it either graduates (the student's mean rollout accuracy on it reaches half or more) or is FIFO-evicted under finite capacity, amplifying BCQ and NCQ inside the student's current zone of proximal development. We post-train Qwen3.5 students at four scales (0.8B-9B) as VLMs with a 27B teacher and evaluate them on a 31-benchmark suite (16 VLM, 10 LLM, 5 Video); ZPPO outperforms off/on-policy distillation and GRPO, with the largest gains at the smallest scale.

cs.CL

版本更新 · v4arXiv:2609.36672

Human-inspired, Task-Dimension-Guided Exploration for Efficient Learning in High Dimensions

Fanyu Zhu, Jiahui An, Ni Ji

原文摘要

Efficient exploration in high-dimensional decision spaces remains a central challenge for decision-making systems. Humans, in contrast, can navigate large decision spaces with remarkable efficiency. Recent behavioral studies suggest that humans reduce dimensionality in large decision spaces by probing candidate feature dimensions, identifying reward-relevant ones, and restricting the effective decision space. Inspired by this mechanism, we propose TDGE (Task-Dimension-Guided Exploration), a human-inspired, model-agnostic algorithm with an automatically constructed task-dimension--feature--item hierarchy. TDGE follows a top-down exploration strategy: it first selects task-relevant feature dimensions, then identifies informative features within those dimensions, and finally recommends concrete items based on the selected features. Experiments on MovieLens-20M, LastFM, and Amazon recommendation datasets show that TDGE substantially improves exploration efficiency and cold-start adaptation over baseline algorithms. Comparisons with other structured algorithms and ablation studies attribute these gains to TDGE's hierarchical structure and semantic feature-space exploration, with robust results across clustering methods and hierarchy depths. Recommendation-trajectory visualizations also show exploration patterns similar to human dimension-guided behavior.

cs.LG

新论文 · v1arXiv:2611.20773

When Does Interaction Help? Toward a Reynolds Number for Collective Intelligence

Xuening Wu, Haode Yang, Yanlan Kang, Shenqin Yin

原文摘要

Multi-agent systems combine distributed information through interaction, which can improve performance but also spread errors or suppress useful differences. Predicting when interaction helps is central to understanding and designing collective intelligence. We investigate whether a few dimensionless parameters can predict gains over explicitly defined non-communicating baselines across group sizes, network structures, and tasks, toward a Reynolds-number-like description of collective intelligence. Using distributed classification tasks, stochastic tracking networks, and neural cellular automata, we combine analytical and empirical evidence to examine information sufficiency, dynamical applicability, and finite-sample prediction. We construct inference systems with bounded random variables and matching specified joint moments up to total degree three, yet opposite interaction gains under a fixed readout. These moments alone cannot generally determine the gain sign. Tracking networks likewise exhibit gain reversals despite matched candidate parameters when task power occupies different spectral modes. Under concentrated task spectra and aligned observation noise, however, we derive a low-dimensional similarity relation with an explicit zero-delay error bound and test it across network sizes and topologies under frozen protocols. The relation does not directly apply to the neural cellular systems, where a separate moment-based predictor underperforms direct gain measurement at matched calibration sample budgets. These results establish conditions and limits for low-dimensional gain prediction, providing a restricted similarity law while leaving a general Reynolds-number-like description of collective intelligence open.

cs.MA

新论文 · v1arXiv:2611.20772

Building Fully On-Device GUI Agents with Mobile NPUs

Zhe Dong, Guanyu Cai, Yuchen Jiang, Jiliang Wang

原文摘要

Mobile GUI agents are increasingly used to automate everyday smartphone tasks, but most existing systems still rely on cloud-hosted models or remote inference, introducing repeated communication latency, privacy exposure, and dependence on network availability. Modern smartphones integrate neural processing units (NPUs) that provide an attractive substrate for fully on-device LLM inference. However, directly deploying conventional GUI-agent pipelines on mobile NPUs is inefficient. GUI-agent inference is fundamentally a long-input, short-output workload, yet conventional agents spend substantial time on autoregressive reasoning and action generation, while mobile NPUs provide much higher throughput for prompt prefill than token-by-token decoding. Meanwhile, dynamically varying GUI-agent contexts conflict with the fixed context profiles required by statically compiled NPU runtimes. We present DroidPrefill, a prefill-centric inference architecture that reshapes the GUI-agent workload around mobile-NPU execution. DroidPrefill combines Prefill-Centric Inference to reduce unnecessary autoregressive decoding with Efficient Context Management to preserve decision-critical information within a fixed compiled context budget. Experiments on commercial smartphones show that, in controlled comparisons with the autoregressive T3A+ baseline, DroidPrefill reduces per-step inference latency by up to 85.0%, while improving task success rate by up to 24.6%.

cs.AR

新论文 · v1arXiv:2611.20771

Nemotron 3.5 Bolt E4B 4GB: An Open and Efficient Small Language Model for Fast Conversational Agents

NVIDIA, :, Adi Margolin, Aditya Malte, Aishwarya Chhabra, Alex Hodson, Ameya Sunil Mahabaleshwarkar, Ambrish Dantrey 等 113 位作者

原文摘要

We introduce Nemotron 3.5 Bolt E4B 4GB, an open and efficient small language model further post-trained from Gemma-4-E4B-it and optimized for speed and intelligence per GB of memory used for fast conversational agents. The model carries forward Gemma's strong conversational capabilities and further improves multi-turn tool use, instruction following, and input grounding for fast, local voice/device assistants, interactive personas, and embedded/Edge interfaces. The model combines natural conversation with multi-turn agentic capabilities to power voice/device assistants, interactive characters and robotic/IoT interfaces running on GeForce RTX, Jetson, and RTX/DGX Spark. Experiments on 21 benchmarks show that Nemotron 3.5 Bolt E4B 4GB achieves the highest average accuracy while generating the fewest output tokens in both reasoning and non-reasoning modes, compared with Gemma-4-E4B-it and Qwen3.5-4B. The model is released with a DFlash drafter. With DFlash, Bolt E4B achieves $2.74\times$ the per-user decode speed of inference without speculative decoding at concurrency~1, and $1.21\times$ the generation throughput of Gemma-4-E4B-it with MTP at batch size 1.

cs.CL · cs.AI · cs.LG

新论文 · v1arXiv:2611.20770

Non-Imaging Optical Tactile Sensors for Robotics: Transduction Principles, System Integration, and Robotic Applications

Hao Wu, Yu Feng, Haotian Guo, Huichan Zhao, Changsheng Wu, Jianshu Zhou

原文摘要

Non-imaging optical tactile sensors encode contact through changes in optical intensity, wavelength, spectra, or photon counts, offering diverse architectures for robotic touch. Their usefulness depends on whether contact variables can be reliably distinguished from coupled mechanical and environmental effects, and whether the resulting information can be delivered in time to guide action. This review develops a unified framework linking mechanical load transmission, optical encoding, readout, and tactile inference to robotic task requirements. It examines reflective and waveguide-based devices, micro/nanofibers, fiber gratings, interferometric and resonant structures, and mechanoluminescent composites. Three considerations emerge across these approaches. First, high optical sensitivity supports accurate tactile estimation only when target variables remain identifiable under the relevant contact and calibration conditions. Second, spatial coverage and feedback timing must be assessed at the system level, accounting for multiplexing, acquisition, demodulation, and inference. Third, device characterization, offline recognition, and closed-loop robotic performance provide distinct levels of evidence and require different evaluation criteria. Representative studies illustrate how these considerations shape force estimation, contact localization, slip correction and distributed interaction. This review concludes with priorities for calibration transfer, reliable interfaces, scalable readout, and task-based validation, providing a basis for selecting and developing non-imaging optical tactile architectures for robotics.

cs.RO

新论文 · v1arXiv:2611.20769

Deep Learning for Satellite-Based Mountain Hazard Screening: A Retrospective Failure Analysis on the 2021 Chamoli Rock-Ice Avalanche

Snigdhajyoti Rakshit

原文摘要

The approach utilizes 10 bands from Sentinel-2 imagery, along with terrain information, in a deep learning pipeline to classify potential hazard sources in any given mountainous region. The performance of our model has been evaluated against the data collected during the rock-ice avalanche event in the Chamoli region in the Indian Himalayas on 7th February 2021. U-Net performs segmentation of the detected land surface in the satellite image, detects changes in the identify candidate hazard zones, and then the following steps are performed to calculate the potential energy using the Monte Carlo sampling for steepest descent (D8) flow paths, finding the exposed structures in the downstream locations, and ranking the zones depending on the risk score. In comparison to the simple spectral rule, our approach achieved the higher macro F1 score of 0.648 (as compared to 0.519). But the F1 score in case of rock and soil was low, being in the range of 0.3s. In case of optical change detection in winter, it's only that northeast-facing slope which couldn't be detected, as it was under 73% of shade prior to the incident. The parameters have been set for this particular scene alone and not independently; hence, this is a retrospective test, and this test locates a failure which has occurred earlier. The primary contribution of this paper is locating the pipeline leak.

cs.CV