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

+479新增论文
+361其中 AI 相关
+64已有论文修订
+95新增 AI 导读

arXiv 最近完成扫描:2026-11-16 01:34 · 全站最近入库:2026-11-16 21:56各信源进度 →

统计口径与数据新鲜度

统计 2026-11-16 02:10 — 2026-11-17 02:10(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:2026-11-16 01:34最近发现新增版本:2026-11-16 21:56库内论文最新更新时间:2026-11-14 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。

23,827 已收录论文19,941 AI 相关6,922 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 23,827 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-16 01:33(北京时间)
版本更新 · v2arXiv:2503.06993

CAPT: Class-Aware Prompt Tuning for Federated Long-Tailed Learning with Vision-Language Model

Shihao Hou, Chikai Shang, Jiacheng Yang, Xinyi Shang, Shreyank N. Gowda, Yan Yan, Yang Lu

原文摘要

Effectively addressing the joint effects of non-IID data and long-tailed distribution remains a key challenge in federated learning. While fine-tuning vision-language models (VLMs) like CLIP has shown promise in tackling non-IID scenarios, it often degrades tail-class performance when both heterogeneity and long-tailed distributions are present, even leading to overall performance falling below the zero-shot baseline. To overcome this, we propose Class-Aware Prompt Tuning (CAPT), a novel framework leveraging pre-trained VLMs to effectively learn from non-IID and long-tailed data. Specifically, CAPT introduces a dual-prompt mechanism that synergizes class-aware and general prompts, enabling the model to preserve class-specific information while capturing global trends. Furthermore, to guide the prompts in learning targeted knowledge, we propose a heterogeneity-aware clustering strategy with two key components: similarity-based clusters aggregate class-aware prompts to reinforce tail-specific knowledge, while heterogeneity-based clusters aggregate general prompts to balance diverse representations and promote robust generalization. Extensive experiments demonstrate that CAPT significantly outperforms state-of-the-art methods, e.g., by improving overall accuracy by 6.1% on ImageNet-LT and 8.8% on Places-LT over baselines, while also substantially enhancing tail-class performance across other benchmarks.

cs.LG · cs.CV

新论文 · v1arXiv:2611.03000

Fluid Antenna Array-Enabled Near-Field ISAC: Joint Port Position Optimization and Transmit Beamforming Design

Wangqi Shi, Zhentian Zhang, Hao Jiang, Zaichen Zhang, Kin-Fai Tong

原文摘要

Fluid antenna systems (FASs) have shown that activating a single port from a densely packed antenna array can yield spatial diversity comparable to that of tens of independent antennas, a property that has substantially improved communication reliability and efficiency. The fluid antenna array (FAA) extends this concept to multi-port configurations while avoiding complex multi-dimensional spatial correlation, and offers significant geometrical flexibility under a finite aperture and a limited port budget. These attributes motivate environment-adaptive array geometry for sensing in integrated sensing and communication (ISAC) systems. Although this flexibility enables adaptive array geometries that exploit near-field distance-angle information, improving multi-parameter target estimation under communication and deployment constraints remains challenging. In this paper, we propose a planar FAA-assisted near-field ISAC framework for multiuser communication and target sensing. Specifically, we derive the Fisher information matrix (FIM) and the corresponding Cramér-Rao bound (CRB) for the joint estimation of target range, zenith angle, azimuth angle, and radial velocity. Based on the derived CRB, we jointly optimize the FAA port positions and transmit beamforming to improve sensing accuracy while satisfying communication signal-to-interference-plus-noise ratio (SINR), transmit power, and physical deployment constraints. To solve the resulting non-convex problem, we develop a two-stage alternating optimization algorithm, in which transmit beamforming is optimized through semidefinite relaxation (SDR), while FAA port positions are successively optimized through discrete search and continuous refinement. Simulation results indicate that the proposed algorithm achieves favorable sensing performance while maintaining a reasonable computational cost.

cs.IT

新论文 · v1arXiv:2611.02999

SyncHeC: Markerless Spatio-Temporal Hand-Eye Calibration via Differentiable Rendering and Continuous-Time Joint Optimization

Zhengdong Hong

原文摘要

Markerless hand-eye calibration under continuous motion requires jointly resolving camera geometry and the time offset between images and joint states. We present SyncHeC (Synchronized Hand-Eye Calibration), a dense differentiable-rendering framework that estimates both from a single robot trajectory. Continuous-time forward kinematics makes latency differentiable, metric depth complements mask contours, and an offline planner combines spatial sensitivity with temporal excitation. Alternating updates support Eye-to-Hand, Eye-in-Hand, and multi-camera calibration. On a 7-DoF arm, one $6.5$\,s motion yields $0.21$\,cm translation error against a marker-board reference, $0.34$\,cm touch-test error, and $2.7$\,ms latency error relative to an $87$\,ms reference with $95\%$ uncertainty of $\pm1.5$\,ms. Against CM-Calib on matched recordings, translation and nominal latency errors improve by $3.7\times$ and $2.3\times$. Experiments span seven arms and four depth sensors. Compensating the recovered offset during policy training and deployment increases success by $8.1$ percentage points across six families and six tasks ($1{,}800$ matched pairs; paired $95\%$ confidence interval $[6.3,9.9]$ points).

cs.RO

新论文 · v1arXiv:2611.02996

LLMs and the Organization of Scientific Work

Stefano Bianchini, Georgiana-Ioana Coroama

原文摘要

The advent of large language models (LLMs) represents a technological revolution for science. Their diffusion may reshape not only how scientists conduct research, but also how they organize their work within teams. Here, we develop a conceptual framework in which LLMs expand individual capabilities across a range of research activities, reducing reliance on complementary human inputs within tasks and facilitating a more specialized division of labor among team members. We further propose that teams' history of collaboration may constrain their ability to reorganize in response to the new technological capabilities. Using CRediT contribution statements from approximately 120,000 PLOS ONE articles published between 2018 and 2025, we investigate changes in the division of labor within scientific teams following the release of ChatGPT in late 2022. We document three main findings. First, participation in scientific tasks falls below its previous trajectory, particularly in conceptualization, execution, and writing. Second, researchers become more specialized, exhibiting narrower task portfolios and less overlap in their contributions with co-authors. Third, these changes are concentrated among newly formed teams, whose members have no prior collaboration ties, whereas established teams, whose members have all collaborated previously, exhibit considerably greater organizational inertia. Our findings suggest that, in the wake of LLM diffusion, scientific work is being reorganized around greater individual autonomy and an increasingly specialized division of labor within teams.

cs.CY

新论文 · v1arXiv:2611.02995

ILLRAW: Illumination-Aware RAW Object Detection

Xinzhe Wang, Xiao Liang, Xiaoli Zhang, Kang Ma

原文摘要

Existing RAW-based detectors typically introduce learnable ISP modules to convert RAW inputs into sRGB-like formats to fit off-the-shelf detectors. However, this RAW-to-sRGB paradigm decouples representation learning from the downstream detection task. Crucially, it forces the network to inherit human-centric visual priors while discarding intrinsic RAW advantages, such as linear sensor responses and high-bit-depth, minimally processed scene cues. In this paper, we propose \textbf{ILLRAW}, an \textbf{ILL}umination-aware end-to-end \textbf{RAW} object detection framework. Instead of reconstructing visually plausible sRGB images, ILLRAW directly learns detection-oriented RAW features through a lightweight feature adapter inserted at the input stage of existing detectors. ILLRAW estimates an illumination mask and introduces channel-sensitive embeddings to model Bayer-channel responses. These features are fused by an illumination attention fusion module. Experiments show clear improvements on the diverse AODRaw benchmark and competitive performance on LOD and PascalRAW, with particularly strong results under dark exposure.

cs.CV

新论文 · v1arXiv:2611.02992

MS-DOTT: A Multi-Modal Dataset for Off-Road Traversability Estimation

Shivani Subramanyan, Saksham Sharma, Samahith SA, Ahaan Kotian, Suresh Sundaram

原文摘要

Autonomous robots are gaining traction in off-road environments, driven by their rising demand in agriculture, defense, and search-and-rescue operations. However, these environments are inherently unpredictable and navigation remains a formidable challenge given the complex environmental conditions including challenging terrain and varying illumination. Off-road autonomy requires robust perception across multiple spatial scales and imaging modalities, yet most sensing setups struggle to capture both far-field environmental context and near-field terrain detail simultaneously across various illumination conditions. To facilitate development of robust off-road perception and traversability estimation systems, we propose MS-DOTT (Multi-Sensor Dataset for Off-road Terrain Traversability) a comprehensive multi-modal off-road dataset with long-range 3D sensing and short-range, fine-grained perception to facilitate high-fidelity terrain understanding by combining proprioceptive signals for tracking vehicle dynamics with exteroceptive imaging and pointcloud data. The dataset includes trajectories recorded across nominal lighting, low-light, and no-light conditions with diverse 2D and 3D sensing for both long-range large-scale and short-range high-resolution sensing, as well as 4D RADAR measurements that provide velocity estimates for detected pointclouds. We demonstrate the utility of the dataset by adapting a SOTA multi-sensor fusion method for the purpose of traversability estimation. This dataset will aid the robotics community in advancing research and development of methods for traversability estimation, off-road perception and multi-sensor fusion in unstructured environments.

cs.RO

新论文 · v1arXiv:2611.02990

Same Function, Different CPU: Quantifying Hardware Heterogeneity in Commercial FaaS Platforms

Jonas Heisterberg, Natalie Carl, Trever Schirmer, David Bermbach

原文摘要

Serverless computing, particularly Function-as-a-Service (FaaS), abstracts operational concerns such as scaling, scheduling, and resource management from users. This abstraction, however, conceals a heterogeneous hardware fleet on which function execution environments are scheduled behind a unified function interface. Consequently, otherwise identical invocations may run on different CPU types that differ in performance, a hardware lottery that users can neither see nor control. Since the resulting variability is entangled with other sources, such as noisy neighbors or diurnal effects, its contribution to overall performance variability and how this contribution differs across providers remain unclear. In this paper, we systematically study hardware heterogeneity and its impact on function performance across four widely used commercial FaaS platforms, covering regional, configurational, and temporal differences. Using variance decomposition, we find that CPU heterogeneity alone explains 4 % to 97 % of the performance variance on AWS Lambda and Google Cloud Run Functions. On Azure Functions, in contrast, we identify a platform-level mechanism that produces bimodal execution durations and overshadows CPU effects. We observe that the CPU composition on AWS Lambda shifts within a single week. Additionally, we reveal that on Google Cloud Run Functions and Alibaba Function Compute, the availability of instruction set extensions such as SHA-NI differs between CPU types, widening the performance gap further. Since all platforms bill by duration, performance gaps between CPU types translate directly into cost, with invocations on the slowest CPU type costing up to three times as much as those on the fastest.

cs.DC

版本更新 · v2arXiv:2604.17652

Self-Supervised Super-Resolution for Sentinel-5P Hyperspectral Images

Hyam Omar Ali, Antoine Crosnier, Romain Abraham, Baptiste Combelles, Fabrice Jégou, Bruno Galerne

原文摘要

Sentinel-5P (S5P) plays a critical role in atmospheric monitoring; however, its spatial resolution limits fine-scale analysis. Existing super-resolution (SR) approaches rely on supervised learning with synthetic low-resolution (LR) data, since true high-resolution (HR) data do not exist, limiting their applicability to real observations. We propose a self-supervised hyperspectral SR framework for S5P that enables training without HR ground truth. The method combines Stein's Unbiased Risk Estimator (SURE) with an equivariant imaging constraint, incorporating the S5P degradation operator and noise statistics derived from signal-to-noise ratio (SNR) metadata. We also introduce depthwise separable convolution U-Net architectures designed for efficiency and spectral fidelity. The framework is evaluated in two settings: (i) LR-HR, where synthetic LR data are used for direct comparison with supervised learning, and (ii) GT-SHR, where super-resolved images surpass the native spatial resolution without HR reference. Results across multiple bands show that self-supervised models achieve performance comparable to supervised methods while maintaining strong consistency. Qualitative analysis shows improved spatial detail over bicubic interpolation, and validation with EMIT data confirms that reconstructed structures are physically meaningful. Code is available at https://github.com/hyamomar/Sentinel-5P-Super-Resolution/tree/main/self_supervised

cs.CV

新论文 · v1arXiv:2611.02989

Finding Errors Is Not Enough: Selective Model Escalation for Entity-Targeted Sentiment Analysis

Nishan Chatterjee, Matthew Purver, Antoine Doucet, Senja Pollak

原文摘要

Improving a sentiment classifier with limited resources involves more than simply deciding how much data to collect or which model to train. A labelled collection may lack examples of an important class. Once trained, a classifier may benefit from another model even when that model is weaker on its own. We study these decisions in document-level Entity-Level Sentiment Analysis (ELSA), primarily using South Slavic news. We first compare acquisition orders as collections grow. We then use uncertainty to choose which predictions to revisit, how many to process, and whether an alternative encoder or local LLM can correct them. Finally, we select complete routing policies on development data for held-out evaluation. Minority-prioritised acquisition helps at some very small training sizes, but not consistently thereafter. On SentiCoref, an oracle correcting all errors in the least-confident 10% without harming correct predictions would add 9.14 Macro-F1 points; however the best evaluated action gains only 0.90. Encoder averaging helps on AspectBench and NewsMTSC but harms SentiCoref, where selective LLM assistance is more promising. Both development-selected policies route 30% of records to the encoder-assisted LLM, yielding observed gains of 1.02 points on SentiCoref (interval includes zero) and 1.71 on NewsMTSC, at substantially different inference costs.

cs.CL

新论文 · v1arXiv:2611.02988

Can You Predict the Next Mind? Toward Multi-Turn Multi-Party Mental World Modeling

Jiahao Huang, Zheng Lian

原文摘要

While multimodal foundation models have achieved remarkable progress in physical world modeling, generation, and embodied reasoning, their capacity for dynamic mental world modeling remains underexplored. Existing benchmarks predominantly focus on static mental state perception or short-horizon decision-making, overlooking the dynamic mental trajectories inherent in multi-party, multi-turn interactions. To bridge this gap, we introduce M$^3$WME-Bench, a benchmark designed to evaluate mental world modeling conditioned on the evolving actions of interactants. Specifically, models are tasked with forecasting a target subject's future mental states and behaviors over extended social interactions. In addition to the benchmark suite, we construct M$^3$WME-23K, a dedicated training dataset, and establish a competitive baseline named MindNext-7B. Extensive evaluations across four representative paradigms (general MLLMs, emotion-specialized MLLMs, agentic frameworks, and World Foundation Models) reveal that current systems consistently struggle to anticipate forward-looking mental states. Mechanistic analyses further reveal that existing architectures fundamentally lack forward mental state transition dynamics. Our findings highlight a pivotal challenge in advancing agents from physical to mental world modeling, charting new directions for socially adaptive AI.

cs.HC

新论文 · v1arXiv:2611.03756

Language Models Can Benefit from Domain-Specific Inductive Biases

Damien Teney, Liangze Jiang, Zachary Shinnick, Hemanth Saratchandran, Simon Lucey

原文摘要

Language models are typically trained with a single choice of architecture on diverse domains and data sources. Modern transformers have been extensively refined for such general-purpose learning, but are they equally effective across different subsets of the pretraining data? We investigate this question by training small language models from scratch on representative subsets of pretraining data such as web text, code, and mathematics. We introduce variants of transformers in which standard activation functions are replaced with learned splines tuned on specific datasets, providing a simple way to modify their inductive biases. Using a mix-and-match of architectures and datasets, we evaluate cross-domain transfer, i.e. how well does an architecture tuned for one domain perform on data from another? We find that the tuned transformer variants improve training efficiency or generalization on all seven of our tested datasets, but the gains are often domain-specific. For example, architectures tuned on a natural language dataset such as FineWeb perform well on TinyStories, whereas those tuned on mathematics perform poorly on natural language. We identify several mechanisms underlying these improvements, including wider loss basins and more abundant "lottery tickets". Our results show that different subsets of pretraining data can benefit from different learning mechanisms, highlighting a limitation of the prevailing trend towards a small set of general-purpose architectures and optimizers. This points to an opportunity to improve data efficiency by developing domain-specific learning strategies that retain the ability to train general-purpose models.

cs.LG

新论文 · v1arXiv:2611.02985

Distilling Literary Sentiment from LLM Annotations into an Ordinal BERT Model

Artem Suslov, Takehiro Hashimoto

原文摘要

The scarcity of human-annotated literary corpora limits the development of domain-specific sentiment analysis models. This study investigates whether LLM-generated annotations can supervise smaller models for literary sentiment analysis. We fine-tune five BERT models using ordinal regression on 10,889 literary sentences annotated by Gemma 4 31B on a ten-point sentiment scale. The models are evaluated against 1,923 human-annotated sentences from Hemingway's "The Old Man and the Sea", alongside the teacher model and conventional baselines. BERT-large achieves a Kendall's $τ$ of 0.589, approaching the teacher's 0.604. Although a constant-neutral predictor achieves the lowest overall mean absolute error (MAE), BERT-base outperforms the teacher on non-neutral sentences, achieving an MAE of 0.726 versus 1.116. These findings demonstrate the potential of LLM-generated supervision for developing smaller literary sentiment models, while highlighting the limitations of aggregate error metrics on predominantly neutral datasets.

cs.CL

版本更新 · v2arXiv:2603.06362

Computer vision-based estimation of small arthropod biomass

Mikko Impiö, Philipp M. Rehsen, Jarrett Blair, Cecilie Mielec, Arne J. Beermann, Florian Leese, Toke T. Høye, Jenni Raitoharju

原文摘要

The ability to estimate invertebrate biomass using only images could help scaling up quantitative biodiversity monitoring efforts. Computer vision-based methods have the potential to omit the manual, time-consuming, and destructive process of dry weighing specimens. We present two approaches for dry mass estimation that do not require additional manual effort apart from imaging the specimens: fitting a linear model with novel predictors, automatically calculated by an imaging device, and training a family of end-to-end deep neural networks for the task, using single-view, multi-view, and metadata-aware architectures. We propose using area and sinking speed as predictors. These can be calculated with BIODISCOVER, which is a dual-camera system that captures image sequences of specimens sinking in an ethanol column. For this study, we collected a large dataset of dry mass measurement and image sequence pairs to train and evaluate models. We show that our methods can estimate specimen dry mass even with complex and visually diverse specimen morphologies. Combined with automatic taxonomic classification, our approach achieves a median individual error of 10%-20% and produces group-level dry mass distributions that do not differ significantly from observed distributions. We highlight the importance of choosing appropriate evaluation metrics, and encourage using both percentage errors and absolute errors as metrics, because they measure different properties. We also explore different optimization losses, data augmentation methods, and model architectures for training deep-learning models.

cs.CV

版本更新 · v5arXiv:2609.39373

Dynamic Coded Distributed Computing with Joint File Placement and Reduce Function Reassignment

Chenglin Li, Minquan Cheng, Youlong Wu

原文摘要

Coded distributed computing (CDC) mitigates the Shuffle communication bottleneck through coded multicast, but existing schemes with predetermined worker populations do not directly adapt to worker arrivals and disconnections. To address these dynamics, we propose a dynamic CDC scheme that preserves surviving workers' caches and Reduce assignments while exploiting the storage brought by arriving clusters. The model accommodates general MapReduce tasks under a placement that is partly fixed and partly new. The scheme uses Cartesian products of cluster-wise placement delivery arrays (PDAs) for file placement. A delivery-aware greedy algorithm reassigns abandoned Reduce functions, and a Communication Array allows workers with no Reduce function to serve as coded transmitters. For arbitrary feasible disconnections, the scheme's exact communication load is derived. A converse bound under one-shot communication shows that, in the absence of disconnections, the scheme lies within a factor of 2 of the lower bound under these assumptions when new clusters arrive, and within a factor of 4 otherwise.

cs.IT

新论文 · v1arXiv:2611.02983

Supporting Habit Formation in Pre-Meal Photo Capture: A Longitudinal Study of Behavior Change Techniques

Jun Fang, Yueying Suo, Yuntao Wang, Susu Wang, Xiaoyu Yang, Qiqi Liu, Chuangyu Ding, Yuanchun Shi

原文摘要

Pre-meal photo capture is lightweight only with a smartphone, yet habit formation requires repeatedly initiating and completing the action within the narrow window before the first bite. We examine how supports based on behavior change techniques can shape capture adherence, perceived habit, and support needs over time. Using a participant-informed selection process, we implemented three smartphone supports targeting Prompts/cues, Focus on past success, and Social support. We evaluated them in a five-week field study with 54 participants, spanning support, withdrawal, and reinstatement, followed by interviews and a three-week follow-up. Across 4,945 main meals, 63.9\% were confirmed pre-meal captures, and adherence was broadly maintained across phases. Perceived automaticity and spontaneous remembering improved mainly during initial support. As capture became more familiar, participants combined spontaneous remembering with deliberate strategies, reduced active feature use while continuing to value support availability, and carried the practice forward through accumulated records, familiar tools, and chosen relationships. These findings frame BCT support for pre-meal photo capture as a configurable arrangement. Systems should preserve a predictable capture pathway while adapting when, how, and through whom support remains available.

cs.HC

版本更新 · v5arXiv:2510.25502

TempoPFN: Synthetic Pre-training of Linear RNNs for Zero-shot Time Series Forecasting

Vladyslav Moroshan, Julien Siems, Arber Zela, Timur Carstensen, Frank Hutter

原文摘要

Foundation models for zero-shot time series forecasting face challenges in efficient long-horizon prediction and reproducibility, with existing synthetic-only approaches underperforming on challenging benchmarks. This paper presents TempoPFN, a univariate time series foundation model based on linear Recurrent Neural Networks (RNNs) pre-trained exclusively on synthetic data. The model uses a GatedDeltaProduct architecture with state-weaving for fully parallelizable training across sequence lengths, eliminating the need for windowing or summarization techniques while maintaining robust temporal state-tracking. Our comprehensive synthetic data pipeline unifies diverse generators, including stochastic differential equations, Gaussian processes, and audio synthesis, with novel augmentations. In zero-shot evaluations on the Gift-Eval, fev-bench and Chronos-ZS benchmarks, TempoPFN achieves top-tier competitive performance, surpassing the majority of models trained on real-world data, while leveraging fully parallelizable training and inference. We open-source our complete data generation pipeline and training code, providing a reproducible foundation for future research.

cs.LG · cs.AI · stat.ML

版本更新 · v11arXiv:2504.04430

Foundational Requirements for Artificial General Intelligence: A Falsifiable Framework Based on Signal Prediction

Matej Šprogar

原文摘要

Grounded in the premise that high-level intelligence can emerge from low-level signal processing, we hypothesise a set of requirements necessary for artificial general intelligence. They characterise systems that learn by predicting spatially and temporally structured signals of initially unknown meaning, and draw on principles observed in cognitive neuroscience, from learning from an uninformed state to real-time liveness. To enable empirical testing and rejection of the hypothesis, we introduce a testbed of transparent, reusable tests, one per requirement. Failing a test exposes a violation of a requirement, but passing is not conclusive: a transparent testbed cannot detect systems engineered around its tests, and such systems can pass it without being intelligent. Passing is therefore meaningful only for systems developed without knowledge of the tests, and none has yet been reported. The testbed thus serves as a necessary-condition filter rather than as proof of progress toward general intelligence. Its reference implementation is publicly available.

cs.AI

新论文 · v1arXiv:2611.02981

Utility-Aware Generative Recommendation: Utility-Enriched Semantic Codes and Utility-Guided Decoding

Yang Hu, Duo Wang, Ning Li, Kang Li, Junda She, Chengcheng Guo, Kuo Cai, Bin Liu 等 10 位作者

原文摘要

Generative recommendation has been widely adopted in large-scale industrial recommendation systems, where multimodal item embeddings are quantized into offline semantic identifiers (SIDs) and items are retrieved by autoregressively generating their SIDs. However, conventional SID construction often underrepresents item utility reflected in post-exposure behaviors such as shares, long views, and video completion, making semantically similar items with different utility insufficiently distinguishable while struggling to accommodate rapidly changing utility signals. Moreover, generation likelihood may be misaligned with downstream utility, causing high-utility candidates to be prematurely pruned during beam search before post-generation reranking or training-time reward alignment can effectively compensate. To address these limitations, we propose UAGR, which aligns both SID construction and decoding with downstream multi-objective utility. Our Utility-Enriched Semantic Codes preserve the original semantic information while incorporating item-level utility without increasing SID length or vocabulary size, supporting incremental utility updates with no additional tokenization or sequence-decoding cost. Meanwhile, our Utility-Guided Decoding estimates candidate utility by combining predictions of multiple post-exposure outcomes and incorporates these estimates into beam scoring, reducing the risk of prematurely pruning high-utility paths while enabling flexible utility control without interfering with generative backbone training. Extensive offline experiments and online A/B tests under real-world traffic demonstrate the effectiveness and deployability of UAGR, which has been fully deployed on Kuaishou and increases video watch time by 0.206\% online.

cs.LG

版本更新 · v4arXiv:2602.07543

MSP-LLM: A Structured Large Language Model Framework for Inorganic Material Synthesis Planning

Heewoong Noh, Gyoung S. Na, Namkyeong Lee, Chanyoung Park

原文摘要

Material synthesis planning (MSP) remains a fundamental and underexplored bottleneck in AI-driven materials discovery, as it requires not only identifying suitable precursor materials but also designing coherent sequences of synthesis operations to realize a target material. Although several AI-based approaches have been proposed to address isolated subtasks of MSP, a methodology for solving the entire MSP task has yet to be established. We propose MSP-LLM, a structured LLM-based framework that formulates MSP as a two-stage process composed of two constituent subproblems: precursor prediction (PP) and synthesis operation prediction (SOP). Our approach introduces a discrete material class as an intermediate decision variable that organizes both tasks into a chemically consistent decision chain. For SOP, we further incorporate hierarchical precursor types as synthesis-relevant inductive biases and employ precursor constraint factorization that preserves precursor-related information in the autoregressive decoding state. Extensive experiments show that MSP-LLM consistently outperforms existing methods on both PP and SOP, as well as on the MSP task, demonstrating an effective and scalable framework with practical potential for autonomous MSP.

cs.AI · cond-mat.mtrl-sci

版本更新 · v2arXiv:2609.08789

Silent Revision: Measuring Undisclosed Change in the Safety Frameworks of Frontier AI Developers

Louis Yiven Zhu

原文摘要

Frontier AI developers publish safety frameworks that commit them to evidencing whether their models are dangerous. The European Union and California now treat these documents as instruments of accountability and impose duties on their revision. Neither requires the developer's account to identify each change to a commitment. We introduce the silent revision rate, the share of material changes to a framework's commitments that the developer's published account does not identify, and we release the versioned, hash-pinned corpus needed to compute it. The corpus holds every public version of the safety frameworks of the twelve developers that have published one, with each provider's changelog, redline or announcement. We trace 710 commitment instances across twelve consecutive version pairs, code them against a frozen codebook, and check the coding with three independent re-runs, a blind re-coding of every disputed row and a second coding of a stratified sample. First, 64% of material changes (95% CI 59 to 69) are silent under a strict standard and 30% under a lenient one. Second, the form of the account matters less than its granularity. Narrative and itemised accounts differ by eight points that eight pairs cannot separate from chance, whereas the rate falls with the number of discrete items a provider publishes. Third, 75% of traced changes weaken or remove a commitment, and weakenings are silent more often than strengthenings, at 70% against 52%. A justification explains why a framework changed and an enumeration states what changed, and only the latter makes revision auditable. We argue that publication duties should carry an enumeration duty, which one provider already meets, voluntarily and incompletely.

cs.CY · cs.AI · cs.SE

版本更新 · v2arXiv:2604.21595

A Kernel Nonconformity Score for Multivariate Conformal Prediction

Louis Meyer, Wenkai Xu

原文摘要

Multivariate conformal prediction requires nonconformity scores that compress residual vectors into scalars while preserving certain implicit geometric structure of the residual distribution. We introduce a Multivariate Kernel Score (MKS) that produces prediction regions that explicitly adapt to this geometry. We show that the proposed score resembles the Gaussian process posterior variance, unifying Bayesian uncertainty quantification with the coverage guarantees of frequentist-type. Moreover, the MKS can be decomposed into an anisotropic Maximum Mean Discrepancy (MMD) that interpolates between kernel density estimation and covariance-weighted distance. We prove finite-sample coverage guarantees and establish convergence rates that depend on the effective rank of the kernel-based covariance operator rather than the ambient dimension, enabling dimension-free adaptation. On regression tasks, the MKS reduces the volume of prediction regions significantly, compared to ellipsoidal baselines while maintaining nominal coverage, with larger gains at higher dimensions and tighter coverage levels.

stat.ML · cs.LG

新论文 · v1arXiv:2611.02974

Memory Admission and Result-Retention Sizing in an RV32IMC FPGA Soft Processor

Yihang Dai, Shuyu Xu, Yinfeng Zhou, Chenglei Peng

原文摘要

In FPGA soft processors with shared backing block RAM, memory admission can stall a multiplication after its product has been prepared. We study how admission timing changes result retention in PaceRV, our board-tested five-stage RV32IMC core, using fixed binaries and matched dynamic instructions. Offering data requests one edge earlier reduces geometric-mean Embench cycles by 18.2% and retention opportunities from 375,063 to 4,373. Admission and retention cycle gains are conditional and non-additive. Under slower reads, a 99% capture target selects an 18-edge validity window with registered requests and 28 edges with early requests, although no multiplication matched within both measurement intervals waits longer. A predeclared sizing rule predicts opportunity recovery exactly in all finite tests, including kernels excluded from model development. Independent timing analysis closely estimates execution-cycle savings. The results support choosing admission first, then deciding whether to extend validity and sizing it against an explicit capture or missed-opportunity objective.

cs.AR

新论文 · v1arXiv:2611.02971

Propositional Dynamic Logic with Program Costs

Krishna Manoorkar, Wolfgang Poiger

原文摘要

We introduce $\mathbb{N}\mathtt{PDL}$, a resource-aware variant of Propositional Dynamic Logic ($\mathtt{PDL}$) capable of accounting for program costs measured by natural numbers (such as time or space in the sense of computability theory). In this framework, programs correspond to transition systems weighted by extended natural numbers, and the syntax of our logic includes formulas expressing which properties are reachable by program executions within specific cost constraints. First, we explain how the standard program constructs of non-deterministic choice, sequential composition, iteration, and tests are interpreted in this weighted setting and establish the validity of their corresponding reduction axioms. Using an appropriate adaptation of the (Fischer-Ladner) filtration technique, we proceed to prove decidability and completeness for $\mathbb{N}\mathtt{PDL}$ as our main results. We furthermore provide upper and lower complexity bounds for this logic using translations between standard $\mathtt{PDL}$ and $\mathbb{N}\mathtt{PDL}$.

cs.LO · math.LO

新论文 · v1arXiv:2611.02969

Revisiting (Bijective) Burrows-Wheeler Transform Space-Time Trade-Off Constructions

Dominik Köppl, Tomohiro I

原文摘要

The bijective Burrows-Wheeler transform (BBWT) replaces the delimiter of the ordinary BWT by a Lyndon factorization and sorts factor conjugates in $ω$-order. It admits quadratic-time constant-workspace algorithms and linear-time linear-space algorithms. We extend the small-space BWT trade-off of Crochemore et al. [JDA,'15] to the BBWT and accelerate both constructions. For $τ\in [3..n]$, let $σ_τ\le\min(σ,τ+1)$ be the maximum construction-batch alphabet size. We construct either transform in $ O\!\left( \frac{n^2}τ \frac{\logτ}{\log\logτ} \right) $ time and $O(τσ_τ)$ additional space, assuming $τσ_τ=O(n)$. This improves the transferred $O((n^2/τ)\logτ)$ bound by a $\log\logτ$ factor. If $σ_τ\in w^{O(1)}$, where $w\inΩ(\log n)$ is the word size, the time improves to $ O\!\left( \frac{n^2}τ + \frac{n\logτ}{\log\logτ} \right). $ For inversion, the same bounds hold with the global alphabet size $σ$ in place of $σ_τ$.

cs.DS

新论文 · v1arXiv:2611.02968

ReContact: Reconstruct and Retarget Human Contacts from Video for Contact-Rich Humanoid Manipulation

Zhongkai Gu, Yang Liu, Wei Zhu, Mitsuhiro Hayashibe

原文摘要

Contact transitions allow humanoid robots to redistribute object support across the whole body, enabling manipulation beyond hand-only grasps. Human videos provide a scalable source of such transitions for robots to learn from. However, body occlusion disrupts object pose tracking and corrupts retargeted motion. Moreover, accurately reproducing the object trajectory does not ensure that the robot applies stabilizing forces at the intended contacts, risking slips or drops in later phases. To address these challenges, we propose ReContact, which reconstructs and retargets human-object interactions through an object-centric contact plan. The plan specifies contact phases, contact pairs, anchors, pressing directions, and sticking or sliding modes. Contact priors stabilize object pose reconstruction through occlusion. A residual policy over a frozen whole-body tracker is rewarded for making contact at planned anchors with forces aligned to the specified directions. In simulation, on a thigh-supported regrasp task, ReContact achieves the planned two-palm bottom hold in 83.5 of episodes over three training seeds, versus 0.16% for the best adapted baseline. One policy trained on 247 references re-solved for new placements raises in-workspace tube-transfer success from 20.4% (source reference only) to 97.8%. On a real Unitree G1, policies execute seven contact-rich manipulation and two loco-manipulation tasks.

cs.RO

新论文 · v1arXiv:2611.02966

SHq: Learnable Grids for LLM Weight Quantization

Michael Helcig, Earl Killian, Dan Alistarh

原文摘要

Block-scaled lookup-table (LUT) formats, which represent weights as indices into a small value set (the grid), have become standard for low-bit LLM quantization, ranging from NF4 and BOF4 to the 3-bit LUT instruction in NVIDIA Rubin. While these formats leave the grid definition flexible, existing implementations rely on fixed grids derived either from numerical data types (INT, FP) or assumed weight distributions (NF, BOF). We demonstrate that no single fixed grid is universally optimal; the best fixed grid varies with bit-width: BOFS at 2 bits, NF at 4 and 5 bits, and INT at 6 bits. We introduce SHq, a closed-form family of grids parameterized by two critical degrees of freedom: a companding strength α that interpolates level spacing from uniform to floating-point-like, and an offset c that introduces asymmetry. This offset allows the sign bit of the FP8 block scale, otherwise unused, to select between the grid and its mirror image, up to doubling the number of distinct values available per block. SHq modifies only the lookup table contents and serves as a drop-in replacement for the grid in any such format. Across five models ranging from 3B to 35B parameters, including a mixture-of-experts architecture, evaluated at 2-6 bits using RTN and GPTQ, SHq achieves the lowest KL divergence in every setting at 3 and 4 bits and in most settings at 2, 5 and 6 bits. It outperforms the best fixed grid by up to 34% and remains superior under mixed-precision configurations. At 4.5 bits per weight, the SHq format reduces the KL divergence of NVFP4 by 41-67% while improving downstream accuracy. Grid families incorporating both spacing and offset parameters perform comparably, whereas those lacking an offset exhibit clearly higher error, confirming that these two degrees of freedom drive the observed gains. We therefore propose SHq as the standard grid for block-scaled lookup-table formats.

cs.LG

版本更新 · v3arXiv:2609.17112

Not Another Text Benchmark: Putting the "Visual" Back in Visual Question Answering for Large Video Models

Rwiddhi Chakraborty, Yinong Oliver Wang, Cheng Zhang, Fan Bai, Zhuoran Yu, Michael Kampffmeyer, Yong Jae Lee, Fernando De la Torre 等 9 位作者

原文摘要

Large video models have exhibited impressive performance on a wide range of visual question answering tasks, owing to the rise of powerful, pretrained text and vision encoders. The usefulness of such models have also been demonstrated on a wide range of benchmarks, with an important caveat - the dominant approach in these benchmarks evaluates multiple choice reasoning via text options. This is a natural way to test text-based reasoning in these models, and has led to significant insights regarding model behavior in the community. In this work, we ask a different question - what happens when the evaluation modality is visual, rather than text? We introduce three new vision-centric evaluation benchmarks in temporal frame retrieval, video future prediction, and causal memory distortion, all designed around evaluating visual understanding capabilities in large video models. Our approach complements the existing approaches to evaluate video understanding in frontier models. We show that current frontier models exhibit significant weakness when attempting to reason through visual queries, rather than text. We conclude with an extended analysis section that provides pointers for future improvements in visual understanding for large video models.

cs.CV

新论文 · v1arXiv:2611.02964

One Image Is Enough: Self-Generated Supervision for Personalizing Human Video Diffusion at Test Time

Sadman Rakib Pinon, Adamo Orsini, Yanis Falaki, Linlian Jiang, Jacob Kingi, Yang Wang, Xinxin Zuo

原文摘要

Pose-driven human video diffusion animates a person from a single reference image and a target pose sequence, yet it struggles to preserve identity, rendering the face and garments unfaithfully as the pose departs from it. Personalization is the standard remedy, but its supervision is missing: one image fixes identity at a single pose, while a multi-pose subject video cannot be assumed in practice. Instead, we synthesize the missing supervision. A domain-tuned, pose-conditioned generator produces pseudo-ground-truth frames at selected key poses. An inference-time reference-only sampler steers these candidates toward the subject and keeps the most identity-consistent. A per-subject low-rank adapter then trains on static clips of them, configured from the reference alone, with no subject video, ground truth, or labels. Although lower in fidelity, these synthetic frames supervise more effectively than real ones. Real frames as captured carry motion that erodes the temporal prior, while a static clip conveys appearance alone. On TikTok and UBC Fashion, the adapted model improves the base on every metric, inverts the identity-versus-motion trade-off, and matches or surpasses adaptation on five real frames despite using none. Code and weights are available at https://github.com/SadmanPinon/one-image-is-enough.

cs.CV · cs.GR · cs.LG

新论文 · v1arXiv:2611.02963

Full-Rank Constructions of Explicit Lossless Expanders and Good Small-Set qLTCs

Yali Aharon-Holzman, Michael Ilan, Assaf Reiner, Amnon Ta-Shma

原文摘要

We construct explicit two-sided lossless expanders whose binary adjacency matrices have full rank, and use them to obtain explicit asymptotically good quantum LDPC codes whose specified X- and Z-check matrices have full row rank and satisfy small-set soundness, or confinement, on a linear range. Our lossless expanders are a variant of those of Hsieh, Lubotzky, Mohanty, Reiner, and Zhang, with a free action of a finite 2-group. Under such an action, injectivity and surjectivity of the adjacency matrix are detected on a fixed-size matrix indexed by vertex orbits. Adding a sparse factor to the gadget at the middle vertices of a single type then enforces full rank, with a controlled loss in expansion. To obtain the 2-group action, we observe that reducing the generators of the Lubotzky-Phillips-Sarnak Ramanujan graphs modulo powers of 2, instead of modulo an odd prime, gives Ramanujan Cayley graphs on finite 2-groups. This construction may be of independent interest. Plugging two such expanders into the balanced product of Lin and Hsieh gives the quantum codes, since full rank of the input graphs is inherited by the check matrices. Our construction supplies the code family sought by Hong, Guo, and Lucas. Applying their analysis, within their framework and under their dynamical assumptions, gives quantum memories whose free-energy density is analytic at every positive temperature and whose guaranteed lifetime is $exp(Ω(N))$ at sufficiently low fixed positive temperature, where N is the blocklength. This improves the previously guaranteed stretched-exponential timescale.

quant-ph · cs.CC · cs.DS · math.CO · math.GR

版本更新 · v2arXiv:2607.07181

Three classes of few-weight self-orthogonal projective codes and their applications for LCD codes

Yue Huang, Zhonghao Liang, Chenlu Jia, Yongkang Wan, Qunying Liao

原文摘要

Since self-orthogonal codes, few-weight codes and linear complementary dual codes(LCD codes, for short) have nice applications in coding theory and cryptography, they have received continuous attention. In 2024, by introducing the notion of the augment code, Heng et al.[16] constructed several classes of few-weight self-orthogonal codes basing on defining sets, which are introduced by Ding et al.[7] in 2007. In this manuscript, for two classes of defining sets, we consider the corresponding augmented codes, construct three classes of four-weight self-orthogonal projective codes and determine the parameters of their dual codes. And then, we construct three classes of LCD codes. In particular, we prove that there exists a class of these LCD codes whose dual codes are almost optimal LCD codes according to the sphere packing bound.

cs.IT