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

+0新增论文
+0其中 AI 相关
+0已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-22 21:02 · 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 2026-11-22 00:51 — 2026-11-23 00:51(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:2026-11-22 21:02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 23,345 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-22 21:00(北京时间)
版本更新 · v3arXiv:2607.12304

What Does a Temporal Benchmark Score Measure? Decomposing Channel Use in Video VLM Evaluation

Farrukh Rahman

原文摘要

A score on a temporal video question answering benchmark is meant to measure that a model has temporal understanding, but it conflates two questions. (1) is the question even temporal, does it need several frames and their order? And (2) when it does, does the model recover the order from the pixels, or read it off the positional encoding (RoPE)? Most of a temporal score answers neither, since a single frame and answer priors often carry it, and the field's validity checks, frame-shuffle sensitivity and multi-frame gain, speak only to (1). For (2) we score a reversal conflict, frames reversed while RoPE stays forward, two ways: label-free, as the accuracy lost against the forward labels, and against reversed labels wherever labels transform deterministically under time reversal. This separates a model that follows the reversed content from one merely disrupted. The two readings dissociate two similarly scoring models at both model scales. Qwen3-VL answers the reversed event it actually sees, reading the visual sequence. Molmo2 correctly reports a backwards video when positions agree, and flipping only the RoPE indices flips its prediction to the forward events, i.e. it reads positions. We call these regimes visual-sequence-dominant and position-dominant. Controls show reversed content is processable whenever the channels agree, and that both families share a bias toward describing events as running forward (arrow-of-time prior) that, being similar in strength, cannot by itself account for the dissociation. The distinction matters because the two channels fail on opposite inputs: two models with the same score are not interchangeable.

cs.CV · cs.LG

版本更新 · v3arXiv:2605.26104

EVIDENT: Routing MLLM Adaptation through Entity-Grounded Visual Evidence for Cross-Domain Video Temporal Grounding

Geo Ahn, Jiwook Han, Youngrae Kim, Joonseok Lee, Jinwoo Choi

原文摘要

Fine-tuning MLLMs for Video Temporal Grounding (VTG) often improves in-domain performance but degrades sharply under domain shift. In this work, we find that this failure is primarily driven not just by unseen query concepts, but by visual domain shift, which prevents the model from coupling its learned temporal localization knowledge with its inherent entity-attention capability. To address this, we introduce EVIDENT, a parameter-efficient adaptation framework that anchors temporal grounding in the inherent entity-attention of pre-trained MLLMs by routing VTG adaptation through explicit visual entity evidence. EVIDENT consists of three components: (i) an Entity Bottleneck Adapter that transforms dense visual tokens into compact entity-level slots, (ii) an Entity-Binding Distillation loss that instills objectness priors into the semantically unstructured MLLM visual space, guiding each slot to bind to a coherent entity, and (iii) an Entity-to-eVidence gating mechanism that leverages the captured entities as evidence, steering the model to localize moments containing query-relevant entities. Together, these components enable VTG fine-tuning to rely on entity-grounded evidence rather than brittle dataset shortcuts. Experiments on cross-domain VTG benchmarks show that EVIDENT consistently improves out-of-domain robustness while preserving competitive in-domain performance with modest parameter overhead. These results suggest that entity-level grounding is an effective inductive bias for generalizable temporal localization.

cs.CV

版本更新 · v2arXiv:2505.22106

AudioTurbo: Fast Text-to-Audio Generation with Rectified Diffusion

Junqi Zhao, Jinheng Zhao, Haohe Liu, Yun Chen, Lu Han, Haoyang Zhang, Xubo Liu, Mark Plumbley 等 9 位作者

原文摘要

Diffusion models have significantly improved the quality and diversity of text-to-audio (TTA) generation but are hindered by slow inference speed. Recent flow-based approaches have been explored for accelerating TTA generation. However, these approaches typically rely on dedicated flow-based generative frameworks and do not directly exploit existing pre-trained diffusion-based TTA models. To address this limitation while leveraging the strong generative capabilities of pre-trained diffusion models, we integrate rectified diffusion into these models to improve TTA generation efficiency. Specifically, we propose AudioTurbo, which learns first-order approximate ordinary differential equation (ODE) paths from deterministic noise-sample pairs generated by a pre-trained TTA model. Experiments show that our approach outperforms a flow-matching-based acceleration model in both generation quality and inference speed.

cs.SD · cs.AI · eess.AS

版本更新 · v4arXiv:2512.05391

LoC-Path: Learning to Compress for Pathology Multimodal Large Language Models

Qingqiao Hu, Weimin Lyu, Meilong Xu, Xiaoling Hu, Kehan Qi, Saumya Gupta, Jiawei Zhou, Chao Chen

原文摘要

Whole Slide Image (WSI) multimodal large language models (MLLMs) are difficult to build and deploy because gigapixel slides induce thousands of highly redundant visual tokens that must be learned from sparse slide-level language supervision. Existing slide-level pathology MLLMs typically combine heavy slide-level encoders with long visual prefixes, making development and deployment expensive under limited computational resources. We revisit this regime and show that WSI tile features are highly redundant at both global and local scales, while task-relevant evidence is sparse and query-dependent. We therefore introduce LoC-Path, a resource-efficient slide-level MLLM that compresses before fusion. LoC-Path uses a sparse token merger and a masked-autoencoder (MAE)-pretrained resampler to replace expensive slide-level encoding with a compact latent interface. A token importance scorer selects the most relevant latents and cross-attention routing adapters fuse them into a few LLM decoder layers. This design lowers both multimodal tuning cost and inference-time latency/memory by avoiding heavy slide-level encoding and long visual prefixes. Extensive experiments show that the inference TFLOPs of LoC-Path are 97% lower and inference memory is also 39% lower. LoC-Path enables efficient end-to-end development and deployment under limited computational resources while remaining competitive with prior slide-level MLLMs. Code is available at https://github.com/WinstonHuTiger/loc_path.

cs.CV

版本更新 · v3arXiv:2609.09152

Silver Rate Is (Almost) Optimal for Gradient Descent

Yuhan Ye, Kaizhao Liu

原文摘要

Gradient descent (GD) is one of the most fundamental algorithms in optimization. Yet even for smooth convex objectives, the optimal rate among predetermined nonnegative stepsize schedules has remained unresolved. Altschuler and Parrilo [Mathematical Programming, 2025] showed that their recursive silver stepsize schedule achieves an $O(n^{-p_{\mathrm{sil}}})$ rate, where $p_{\mathrm{sil}}=\log_2(1+\sqrt2)\approx1.2716$, improving upon the classical $O(n^{-1})$ rate of constant-step GD. Although the optimality of $p_{\mathrm{sil}}$ has been well supported by numerical evidence for small $n$, a matching lower bound has remained missing. In this paper, we prove an $Ω(n^{-p_{\mathrm{sil}}-o(1)})$ lower bound for GD, showing that $p_{\mathrm{sil}}$ is the optimal convergence exponent. Additionally, we prove an anytime lower bound that matches the exponent $2p_{\mathrm{sil}}/(1+p_{\mathrm{sil}})\approx1.1195$ of the known upper bound [Zhang, Lee, Du, and Chen, COLT 2025]. Building on the approaches of [Ma and Chen, 2026; Jung, Cho, and Yun, 2026], we select specific iterations as checkpoints at which the GD trajectory explores a new coordinate, and we couple these coordinates using a chain of local functions. Our main technical contribution is a local hard function that bends the GD trajectory between successive checkpoints, resulting in a sharpened lower bound. For each prescribed stepsize schedule, we maximize this bound over admissible checkpoint sets. A recursive analysis of this maximum then yields a lower bound matching the silver exponent for every such schedule.

math.OC · cs.LG

版本更新 · v2arXiv:2611.23788

Geometry-bridge flow matching for blind rigid-motion correction in head cone-beam CT

Sungho Yun, Seungryong Cho

原文摘要

Rigid motion of a patient head during head cone-beam CT (CBCT) scan can degrade image quality and complicate recovery of accurate anatomical structures from the measured projections. We present a 3D flow-matching prior approach that learns the image changes induced by motion. Its geometry-bridge connects motion-corrupted and motion-free Feldkamp-Davis-Kress (FDK) reconstructed images of the same source volume through progressively reduced simulated motion. The forward projector's geometry derivative supplies analytic velocity targets along this path. A plug-and-play loop encompasses the learned image updates together with spline-based pose estimation, conjugate-gradient data consistency, and image total-variation regularization. Both the refined image volume and the FDK reconstructed image volume at the estimated poses are evaluated. On simulated 360-view scans from 30 held-out CQ500 patients with nominal per-component motion amplitudes within 10 mm / 10 degrees, the framework achieved state-of-the-art image reconstruction quality and motion accuracy in comparison with other methos such as learned autofocus and diffusion-based joint reconstruction. Mean peak signal-to-noise ratio (PSNR) reached 37.1 dB and mean detector-plane reprojection error (RPE) was 0.27 mm. With matched endpoints at the inference settings, the geometry bridge improved PSNR by 1.6 dB and reduced RPE by 33% relative to image-linear training. With 50 prior updates and coarse-to-fine motion estimation, total processing time averaged 9.1 minutes per patient on an RTX A6000.

physics.med-ph · cs.CL · cs.CV

新论文 · v1arXiv:2611.25515

When and why can integral constraints improve the accuracy of physics-informed neural networks?

Myeong-Su Lee

原文摘要

Adding constraints obtained by integrating governing equations to the loss has been reported to improve the accuracy of physics-informed neural networks (PINNs) in several applications, but mathematical understanding of these benefits remains limited. For several partial differential equations (PDEs), we derive stability estimates for losses with and without integral constraints. The comparison reveals that in certain parameter regimes, the standard loss has very weak control over specific error components, and that the integral constraints restore uniform control of these components. This explains when and why the constraints improve accuracy. Finally, numerical experiments support this interpretation. In parameter regimes where the standard loss is predicted to have weak control over those components, the two losses indeed show clear differences in accuracy. Conversely, in regimes where the standard loss retains sufficient control, both losses yield similar accuracy.

math.NA · cs.LG · math.AP

新论文 · v1arXiv:2611.25514

Sequential Probability Assignment against Smoothed Adversaries with Unknown Base Measure

Ziyi Liu, Daniel M. Roy

原文摘要

Smoothed online learning has recently been studied as a way to bypass hardness results for the fully adversarial setting, which can be overly pessimistic. In this framework, the adversary is constrained to generate contexts from distributions having a bounded density with respect to some fixed base measure $μ$. Prior work makes the strong assumption that $μ$ is known to the learner, with notable exceptions including Block et al. (2024) and Blanchard (2025). In this paper, we study sequential probability assignment (a.k.a., online learning with log loss) with smooth, well-specified data in the more general setting where $μ$ is \emph{unknown}, without the Lipschitz condition imposed in the work of Block et al. (2024) and Blanchard (2025). We prove regret upper bounds against both adaptive and oblivious smoothed adversaries: the adaptive bound is controlled by empirical $\ell_\infty$ Hellinger entropy, while the oblivious bound improves this dependence to empirical $\ell_2$ Hellinger entropy, the notion that has been shown to characterize the complexity of learning with i.i.d. data (Bilodeau et al., 2023). Both bounds are obtained through algorithms based on truncated empirical Hellinger covers. We complement these results with a general lower bound, showing that our upper bounds are essentially tight for some natural classes, which implies a separation in the difficulty of smoothed online learning between regimes where $μ$ is known and where it is unknown.

cs.LG · stat.ML

新论文 · v1arXiv:2611.25513

Mesh-Free Keypoint Learning for Category-Level 6D Object Pose Estimation

Junghoon Seo, Hakjin Lee, Jaehoon Sim

原文摘要

Category-level 6D pose estimation relates partial RGB-D observations of unseen instances to a canonical frame shared across a category. Object-aware reconstruction helps adaptive keypoints retain the geometry of the visible surface. Common reconstruction targets select observed points by their distance to a registered instance surface. We present MORSEPose, which derives this support directly from standard pose and metric-size annotations. We map observed 3D points to normalized object coordinates and retain those inside a tolerancedilated pose-and-size envelope. This construction is consistent across camera frames and metric scales. When the annotated extent bounds the object surface, the envelope conservatively contains the support of a conventional surface-distance selector. We integrate this target with multi-level RGB-D features, adaptive keypoints, and learnable category context. Experiments on REAL275 and HouseCat6D show that the envelope closely matches surface-based supervision at both point and task levels, remains stable under boundary perturbations and viewpoint changes, and supports competitive category-level pose accuracy across benchmarks.

cs.CV · cs.RO

版本更新 · v2arXiv:2607.14397

Model-Informed Joint Material-Structural Optimization of Hard-Magnetic Soft Materials

Ian Galloway, Prashant K. Jha

原文摘要

This work develops a model-informed framework for predictive analysis and optimal design of hard-magnetic soft materials (hMSMs), which undergo contact-free, field-driven deformation. Accurate prediction requires effective structure-property relations, while optimal design requires coordinated control of structural density, magnetic particle distribution, and remanent magnetization direction. First, classical rigid-inclusion, Hill self-consistent, and constrained-kinematics models are placed in a unified effective shear-modulus framework for particle-filled elastomers. Six particle-fraction-dependent relations and one default control relation are combined with three strain-energy density functions to obtain 21 constitutive models. The results show that the strain-energy form has a relatively small effect for the actuation problems considered, whereas the effective shear-modulus relation can significantly affect deformation when magnetic material overlaps with substantially deforming regions. Experimental stress-strain data are used to select a representative shear-modulus relation, with the Mooney relation giving the lowest errors for the dataset and metrics considered. Second, the selected model is incorporated into a joint material-structural optimization formulation with separate fields for structural density, magnetic particle volume fraction, and remanent magnetization direction. The particle-fraction field affects both effective stiffness and magnetic loading, while the applied field is prescribed. Rotational, translational, restorative, and three-dimensional morphing examples demonstrate different active design fields, objectives, and single- or multi-load-case formulations, producing non-intuitive hMSM designs with prescribed deformation responses. The framework is implemented in the open-source CEADpx/MatTO repository.

cs.CE

新论文 · v1arXiv:2611.25510

Life-IQA+: Boosting Blind Image Quality Assessment via GCN-enhanced Layer Interaction and Flexible Hierarchical Expert Specialization

Long Tang, Guoquan Zheng, Huiyu Duan, Jie Hao, Yucheng Zhu, Liang Yuan, Xiongkuo Min, Guangtao Zhai

原文摘要

Blind image quality assessment (BIQA) is essential for evaluating and optimizing visual experience. Existing BIQA methods typically aggregate features from multiple backbone stages without accounting for their unequal contributions to quality prediction. Furthermore, uniformly supervising all experts may constrain their flexibility and limit the diversity of the learned quality representations. To address these issues, we present Life-IQA+, a quality representation decoding framework that combines graph convolutional network (GCN) enhanced layer interaction and flexible hierarchical expert specialization. Specifically, the layer interaction module constructs quality queries from GCN-refined deepest-stage features and uses penultimate-stage features as keys and values, thereby enabling cross-layer representation interaction via the cross-attention mechanism. Building on these representations, we further introduce a flexible hierarchical mixture-of-experts (MoE) architecture comprising a densely activated shared expert layer and a sparsely activated routed expert layer. To enhance the representation diversity and ability of experts, we apply auxiliary classification supervision to a subset of shared experts using distortion-type labels for synthetic datasets and discretized quality-level labels for authentic datasets. Experiments on synthetic and authentic distortion benchmarks, together with controlled ablations of expert configurations and supervision strategies, show that Life-IQA+ achieves competitive prediction accuracy with a favorable balance between accuracy and computational cost. The code is available at: https://github.com/TANGLONG2/Life-IQA-PlusPlus.

eess.IV · cs.CV

新论文 · v1arXiv:2611.25509

LingSante MedGuard Technical Report

LingSante MedGuard Team

原文摘要

As medical AI is increasingly used for patient consultation and clinical decision support, identifying harmful, non-compliant, or ethically problematic responses is essential for responsible deployment. Assessing these risks requires checking clinical claims against relevant evidence, verifying calculations, and applying medical requirements under their intended conditions. We introduce LingSante MedGuard, a capability-enhanced guardrail that integrates supervision from a medical guardrail teacher model and six specialized medical capability experts into a unified model. The guardrail teacher supervises risk decisions and explanations across medical safety, compliance, and ethics, while the capability experts support medical rule understanding, numerical reasoning, multimodal hallucination detection, agentic safety verification, risk localization, and multimodal entity understanding. We consolidate these sources of supervision through offline supervised fine-tuning on teacher-generated demonstrations, followed by multi-teacher on-policy distillation using feedback on student-generated responses. We further introduce LingSante MedGuardBench, a benchmark comprising 11 risk-detection subsets spanning medical safety, compliance, and ethics. At 4B, 9B, and 27B scales, LingSante MedGuard achieves mean benchmark F1 scores of 82.3%, 82.4%, and 87.4%, respectively. The 4B model also achieves 73.9% mean accuracy across MedHallu and MedHALT-NOTA under retrieval-enabled evaluation. Additional evaluations of the verification components show improvements in evidence-supported consistency, source traceability, and rule-guided rationale quality across model scales. These results support the integration of specialized medical capabilities into a unified guardrail for medical risk assessment and evidence-based medical safety verification.

cs.CR

新论文 · v1arXiv:2611.25508

Persona Lenses: What Makes a Vision-Language Model See Through Its Persona's Eyes?

Yoonho Shin

原文摘要

Vision-language models have been used to simulate where a user looks on a web page. However, a persona in the prompt reaches the simulated gaze only partly. We measure this reach as the difference in one model's gaze between two MatrAIx personas that differ in a single attribute. We introduce persona lenses, which place the persona between the model and the page: a text lens (a two-step prompt in which the model first writes what the persona would look at), a pixel lens (marks on the region matching the persona's interests) and a weight lens (a dwell reward gated to keep the scanpath natural). Using Qwen3-VL (4B and 32B) and Qwen3.5-9B on generated and live pages, we compare the lenses with one another and with scale, the 32B model with no lens. It has been confirmed that the text lens separates the personas strongly but is not robust to profile changes that should not move gaze. The pixel lens separates the personas less but is unaffected by attributes unrelated to gaze. At 4B and 9B, the weight lens separates the personas on interests held out from training, with no extra step on a new page. The 32B model separates the personas with the base prompt alone. Therefore, the choice of a lens depends on cost and model size.

cs.CV · cs.AI · cs.HC

新论文 · v1arXiv:2611.25506

Carry the Fine Print: Budget-Aware Visual Token Compression for Document OCR

Yixuan Huang, ShengFan Zhang, Jiawei Chen

原文摘要

Visual token compression often treats salience or similarity as a proxy for transcription importance. Complete document OCR breaks this assumption: small symbols and repeated entries can be indispensable to the transcript despite being visually inconspicuous. We present Select and Compensate, a two-stage framework built on a simple idea: choosing where to retain a token need not determine all the evidence it carries. Selection combines adaptive encoding resolution with global importance and local contrast to choose which tokens to retain as anchors. Compensation enriches these anchors with nearby discarded content and its relative position, without adding tokens. The OCR backbone remains frozen, and native coordinates and structural tokens are preserved. We evaluate complete page transcription across five backbones on OmniDocBench and document reconstruction on MPDocBench. With half the original visual tokens, our input-adapted configuration reaches 95.42 OmniDocBench Overall on Ovis versus 96.09 for the uncompressed model, and paired evaluations show gains over simply reducing the input resolution on both benchmarks. With the original input grid, our method stays within 0.17 points of the uncompressed model at 75% retention and exceeds a learned global selector by 4.70 points at 25%. On a fixed 200-page Ovis workload at the same 50% visual-token budget, only 3 of its outputs hit the length limit, versus 30-34 for the evaluated RTPrune, DivPrune, and HiPrune adaptations, and it finishes the batch 35-40% sooner. Code will be made publicly available.

cs.CV

版本更新 · v2arXiv:2611.24481

UEmo-D1: An Efficient and Calibrated Decision Model for Unified Multimodal Emotion Recognition

Daiqing Wu, Junqiao Wang, Chengyang Fang, Runhao Zeng, Fei Ma, Sicheng Zhao

原文摘要

Multimodal Emotion Recognition (MER) seeks to identify emotions conveyed through or elicited by media comprising diverse modality compositions. Research in this field has largely evolved along two paradigms: task-specific models and approaches based on Multimodal Large Language Models (MLLMs). Despite their distinct strengths, neither fully reconciles three capabilities essential to human-like emotion understanding: cross-task generalization, efficient decision-making, and receptiveness to multiple plausible interpretations. Inspired by recent advances in decision models, we explore a unified framework that maps multimodal inputs, guided by natural language instructions, directly to distributions over candidate options, addressing these requirements through direct probabilistic decisions. We instantiate this framework in UEmo-D1, the first decision model for MER to our knowledge. To support joint training and systematic evaluation, we also curate UEmo-Data, comprising 977,270 training samples and 41 evaluation subsets across six MER tasks. Extensive experiments demonstrate UEmo-D1's combination of strong recognition performance, low inference latency, and well-calibrated confidence estimates. Complementary analyses further reveal how training strategies and data composition influence these capabilities, providing practical insights into unified MER modeling. Together, our findings establish decision models as a feasible direction toward human-level emotion recognition. Project page: https://github.com/wdqqdw/UEmo-D1.

cs.MM

版本更新 · v2arXiv:2606.09868

SPACE: Source-free Proxy Anchor Concept Erasure for MLLMs

Zhijing Zhang, Jiaqi Ding, Qianshan Wei, Nan Zhou, Jiaqi Li, Yongliang Wu, Tongxin Zhu, Xiaolin Fang

原文摘要

As Multimodal Large Language Models (MLLMs) face growing privacy risks and regulatory constraints, machine unlearning (MU) has emerged as a crucial solution for removing sensitive data while preserving model performance. However, existing MU methods typically rely on visual data of the target concepts, which is often unavailable due to strict data retention policies, thus creating a demand for source-free unlearning approaches that operate without access to the target data. In this work, we propose Source-free Proxy Anchor Concept Erasure (SPACE), the first source-free unlearning framework specialized for MLLMs. SPACE consists of two stages: (1) Text-Guided Proxy Anchor Selection (TPAS), which retrieves semantically aligned proxy anchors from the shared feature space. (2) Dual-Constraint Semantic Isolation (DCSI), which optimizes these anchors to indirectly erase target concepts. DCSI confines updates to the null space of retained knowledge, ensuring structural integrity. We theoretically prove that SPACE strictly bounds the perturbation on retained knowledge and maximizes feature spectral entropy, thereby maintaining the model's performance. Furthermore, extensive experiments across six datasets show that SPACE achieves performance comparable to that of state-of-the-art data-dependent methods, validating its effectiveness in source-free MU scenarios. The source code will be released.

cs.LG · cs.AI

新论文 · v1arXiv:2611.25503

Joint Language Modeling for In-the-Wild Speech-Scene Generation with Distinct Token Representations

Semin Kim, Ju Yeon Kang, Sunghwan Ahn, Seonuk Kim, Nam Soo Kim

原文摘要

Speech-scene generation aims to synthesize intelligible target speech together with diverse environmental sounds as a coherent acoustic scene. The task is considered challenging in two respects. Speech and environmental audio differ substantially in their semantic and acoustic characteristics, yet existing approaches typically model the complete scene using a single mixed acoustic representation. At the same time, the two sounds are closely related in real-world speech scenes, whereas conventional approaches usually rely on simulated scenes mixed from independently recorded sources. We propose a large language model (LLM)-based framework for joint modeling of distinct speech and environment token streams. A source-aware neural audio codec extracts temporally aligned speech and environment tokens from mixed audio, and a pretrained LLM generates both token streams in an interleaved fashion conditioned on the natural-language scene description and speech content. This formulation preserves source-specific representations while modeling the token streams within a shared generative context. Since both streams are extracted directly from mixed recordings, the model can also be adapted to in-the-wild speech scenes without isolated sources. Our experimental results demonstrate that the proposed framework improves speech intelligibility, scene relevance, and audio quality over existing speech-scene generation techniques, with consistent gains from real-world adaptation. We further investigate the effects of distinct token representations and interleaved generation, and analyze whether the environmental sound remains stable while speech is active and how the two streams interact.

eess.AS · cs.AI · cs.LG

新论文 · v1arXiv:2611.25502

Coordinated Use of Large Language Models in Academic Research

Vladislav Marinov

原文摘要

This study proposes a methodological framework for the coordinated use of multiple large language models in academic research. Its central argument is that involving more models does not in itself guarantee more reliable results; what matters is how cognitive and verification operations are organised. The framework is based on seven principles (procedural independence, functional allocation, the diagnostic value of divergence, the diagnostic value of convergence, external verification, proportionality of checking, and human epistemic responsibility) and sets out a non-linear multi-LLM protocol in which independently obtained results are compared, critically examined and, where necessary, verified against external sources. The approach is illustrated through real-world cases drawn from research practice. The study concludes that moving from a single language model to a coordinated multi-LLM process represents a change in the organisation of research, in which the researcher makes the final decisions and bears scholarly responsibility.

cs.AI · cs.CY

新论文 · v1arXiv:2611.25500

PRIO: Pre-trained Representation for Interactive Objects in 4D Worlds

Shixun Huang, Lifu Wei, Yue Chang, Chen Geng, Heng Zhang, Jinhyung Ahn, Caoliwen Wang, Siyuan Chen 等 11 位作者

原文摘要

3D worlds can now be captured, reconstructed, and generated at scale, yet the objects in them stay still: a pressed pillow does not compress, and a pulled door does not swing. Bringing these objects into interactive 4D world models requires a representation of how they can move, deform, and respond physically to user actions in real time. We present PRIO, a pre-trained representation for interactive objects. Given static geometry, PRIO predicts an object-specific configuration space in a single forward pass. This space provides generalized coordinates for Lagrangian mechanics, allowing physical laws and user actions to govern motion. Configurations are linear in these coordinates, allowing the quantities needed by the dynamics to be precomputed. Dynamics then run at interactive rates without further network inference. Pre-trained on diverse 4D trajectories, PRIO learns priors over object configurations that generalize to unseen shapes. It supports meshes, point clouds, and 3D Gaussians, bringing objects in reconstructed and generated scenes into interaction. We demonstrate its use in real-to-sim workflows and robot manipulation. In addition, compared with existing 4D generation methods, PRIO generates dynamics faster and follows user actions more faithfully.

cs.CV

版本更新 · v4arXiv:2512.16767

Make-It-Poseable: Feed-forward Latent Posing Model for 3D Characters

Zhiyang Guo, Ori Zhang, Jax Xiang, Alan Zhao, Zhenxun Yuan, Wengang Zhou, Houqiang Li

原文摘要

Posing 3D characters is a fundamental task in computer graphics. However, existing paradigms, ranging from traditional auto-rigging to recent pose-conditioned generative models, frequently struggle with inaccurate skinning weights, fixed mesh topologies, and poor pose conformance. These challenges have become particularly pronounced with the recent explosion of AI-generated 3D assets, which often exhibit flawed structures and fused geometry. To address these issues, we introduce \textbf{Make-It-Poseable}, a novel feed-forward framework that reformulates character posing as a skinning-free latent-space transformation problem. By decoupling shape deformation from the constraints of fixed mesh connectivity, our method directly operates on compact latent representations to reconstruct characters in target poses. To achieve this, our framework integrates a latent posing transformer for shape manipulation, a dense pose representation for fine-grained control, and an adaptive completion module optimized via a bipartite-matched latent loss to robustly handle topological changes. Extensive experiments demonstrate that our method significantly outperforms existing baselines in posing quality. Furthermore, our design shows promising generalization to diverse morphologies such as quadrupeds in our qualitative tests, and supports various 3D authoring applications such as part replacement and refinement.

cs.CV

版本更新 · v2arXiv:2602.02227

Show, Don't Tell: Morphing Latent Reasoning into Image Generation

Harold Haodong Chen, Xinxiang Yin, Wen-Jie Shu, Hongfei Zhang, Zixin Zhang, Chenfei Liao, Litao Guo, Qifeng Chen 等 9 位作者

原文摘要

Text-to-image (T2I) generation has achieved remarkable progress, yet existing methods often lack the ability to dynamically reason and refine during generation--a hallmark of human creativity. Current reasoning-augmented paradigms most rely on explicit thought processes, where intermediate reasoning is decoded into discrete text at fixed steps with frequent image decoding and re-encoding, leading to inefficiencies, information loss, and cognitive mismatches. To bridge this gap, we introduce LatentMorph, a novel framework that seamlessly integrates implicit latent reasoning into the T2I generation process. At its core, LatentMorph introduces four lightweight components: (i) a condenser for summarizing intermediate generation states into compact visual memory, (ii) a translator for converting latent thoughts into actionable guidance, (iii) a shaper for dynamically steering next image token predictions, and (iv) an RL-trained invoker for adaptively determining when to invoke reasoning. By performing reasoning entirely in continuous latent spaces, LatentMorph avoids the bottlenecks of explicit reasoning and enables more adaptive self-refinement. Extensive experiments demonstrate that LatentMorph (I) enhances the base model Janus-Pro by $16\%$ on GenEval and $25\%$ on T2I-CompBench; (II) outperforms explicit paradigms (e.g., TwiG) by $15\%$ and $11\%$ on abstract reasoning tasks like WISE and IPV-Txt, (III) while reducing inference time by $44\%$ and token consumption by $51\%$; and (IV) exhibits $71\%$ cognitive alignment with human intuition on reasoning invocation.

cs.CV

版本更新 · v2arXiv:2611.23555

MetaphIR: Corpus-Scale Applicability Analysis for LLVM Peephole Optimizations

Hongxu Xu, Xintong Zhou, Chunhao Liao, Chengnian Sun

原文摘要

Peephole optimizations replace short source instruction sequences with more efficient equivalents, typically guarded by preconditions such as requiring a constant operand to be a power of two. Since every new rewrite adds maintenance burden and compile-time cost, compiler developers need to know how often real-world programs both match the source pattern and satisfy its preconditions before committing to an implementation. Existing workflows fall short: structural search over an intermediate representation (IR) corpus of 4.8 million functions ignores preconditions and therefore overcounts, whereas implementing the rewrite and recompiling the corpus yields exact counts only after paying the engineering cost, with each run of the corpus's on-demand recompilation pipeline taking a median of 47 minutes. This delay is especially limiting when generalizing a rewrite, as developers often iterate over multiple variants to trade off coverage against complexity. We introduce MetaphIR, a system pairing a domain-specific language for expressing LLVM peephole optimizations with an indexed search engine over large IR corpora. Using a syntax modeled on LLVM's C++ matchers, MetaphIR patterns augment LLVM IR with holes and Boolean preconditions over matched values. To answer queries at interactive speed, MetaphIR builds compact per-function summaries ahead of time and uses abstract interpretation over those summaries to soundly filter out functions that cannot satisfy a query's constraints. Across twenty recent peephole optimizations from LLVM's InstCombine pass, MetaphIR recovers the exact set of functions where each rewrite applies in isolation, achieving 100% precision and recall. Queries complete in a median of 1.27 seconds---a 22x speedup over a full corpus scan---enabling developers to quickly gauge an optimization's reach and generality prior to implementation.

cs.PL · cs.SE

版本更新 · v2arXiv:2611.23803

CoT-WAM: Collaborative Training of Vision-Language Models and World-Action Models via Plan Latents

Jinhao Zhang, Yushen Zuo, Ke Rui, Weitao Zhou, Haibao Yu, Honglong Tian, Xiaoheng Ma, Jie Mei 等 10 位作者

原文摘要

General robot manipulation requires both semantic competence and physical foresight. World action models (WAMs) acquire an understanding of and predictive capability over the physical world by jointly modeling visual foresight and action distributions. However, existing WAMs still lack semantic competence: they have limited ability to perform open-world scene understanding, semantic reasoning, and long-horizon planning. Although vision-language models (VLMs) provide these capabilities, effectively integrating their semantic priors into WAM prediction and control remains challenging. To address this challenge, we propose CoT-WAM, a jointly trained WAM-VLM framework. CoT-WAM uses a learnable embedding to exchange information between the VLM and the WAM, which is supervised by textual and goal-image signals during training to encode rich intent information. A tailored two-stage training strategy enables joint optimization and more effective collaboration between the VLM and WAM. We evaluate CoT-WAM on six benchmarks covering control performance, zero-shot generalization, and long-horizon planning. CoT-WAM achieves SOTA performance on LIBERO, CALVIN, and RoboCasa-Kitchen, and demonstrates stronger zero-shot generalization on LIBERO-Plus and LIBERO-Pro without fine-tuning. On RoboCasa365, it further improves long-horizon planning over baselines with and without external planners. Real-world experiments further demonstrate the effectiveness of our method in practical deployments. Code is available at: https://github.com/jhz1192/CoT-WAM.git

cs.RO

版本更新 · v3arXiv:2609.27964

Learning from Transients: Linear RNN Scaling Laws in Stationary Prediction

Ziyan Chen, Linhe Zheng, Peilin Liu, Zhongzhu Zhou, Ding-Xuan Zhou

原文摘要

Does the organization of a fixed number of transitions into independent trajectories affect a fixed stationary prediction task? We study a Gaussian-sketched linear recurrent teacher--student model, where independently initialized training trajectories create transients that can excite weak spectral directions. For full-batch gradient descent with stable deterministic learning rates, we establish matching stationary approximation, bias, variance, and total-risk scaling laws from dependent trajectory data. When initialization decays at least as fast spectrally as the innovations, trajectory count and length enter the leading variance rate through their product. With a heavier initialization tail, transient excitation can improve the trajectory-length-dependent finite-time bias and, at fixed transition count, variance, while the stationary oracle shift remains controlled by representation approximation. Under token and compute caps, constant-length trajectories attain the optimal risk scaling order within the theorem-controlled range. Synthetic experiments provide finite-scale evidence in this stylized model.

cs.LG

版本更新 · v3arXiv:2609.27365

Anchor and Perturb: Lazy Agent Remediation by Exploration Injection

Chengxi Zhong, Yongzhe Chang

原文摘要

Anchor and Perturb (AnP) is a lightweight framework that resolves multi-agent coordination failures by decoupling exploratory variance injection from recurrent manifold stability. Existing remediation strategies predominantly alter mixing network architectures or enforce simultaneous exploration across the collective, which inevitably precipitates severe temporal-difference penalties in non-monotonic reward spaces. Specifically, AnP isolates underperforming lazy agents and injects an asymmetric exploratory pulse into targeted coordinates whilst anchoring converged teammates to nominal greedy exploitation. Empirical telemetry benchmarks demonstrate that AnP successfully rescues collapsed joint policies (recovering from a 5% evaluation win rate nadir back to 85%) and facilitates escape from suboptimal coordination plateaus, sustaining peak win rates of 90% without requiring structural network modifications.

cs.MA

版本更新 · v4arXiv:2410.07468

Fast Mixed-Precision Real Evaluation

Artem Yadrov, Aditya Kumar, Pavel Panchekha

原文摘要

Evaluating real numbers to high precision is important in computational science, numerical computing, and in hardware and software design. Efficient algorithms exist for computing arithmetic operations and common elementary functions, but composing these primitives into larger expressions remains a challenge. The typical expression-evaluation algorithm computes all primitives in the expression to a uniform precision, doubling that precision until a sufficiently-accurate result is achieved. This is wasteful: usually only a few operations really need to be performed at high precision, and the bulk of the expression could be computed at a much lower, faster precision. However, no rigorous algorithm currently exists to compute these non-uniform precision assignments, nor to leverage non-uniform precision to reduce the number of primitive computations needed to evaluate an expression. We propose a new algorithm for deriving such rigorous precision assignments and reducing the number of primitive computations needed. This algorithm can leverage results computed at lower precisions to analytically determine a precision for each primitive that will result in a sufficiently-accurate evaluation. It can also use information from constant subexpressions, movability flags, and repeat flags to skip the computation of certain primitives, further reducing the time to evaluate real expressions. Our implementation, Reval, achieves an average speed-up of 3.97x compared to the widely-used Sollya tool, with the speed-up increasing to 9.26x on the most difficult inputs. A detailed evaluation shows that this speed-up results from both assigning lower precisions to the majority of primitives and from computing fewer primitives overall.

cs.MS · math.NA

新论文 · v1arXiv:2611.25489

When Should a Voice Agent Speak? End-of-Turn Detection in Turkish

Ahmet Tuğrul Bayrak, Fatma Nur Korkmaz, Mustafa Sertaç Türkel, Nur Yağmur Tuncer

原文摘要

A voice agent that waits for a fixed silence before replying either interrupts users who pause mid-utterance or responds late to users who have finished. This study examines the decision at the onset of silence, using turn ends and turn-internal pauses extracted from the Real-TurnTurk corpus of Turkish dyadic video calls. A fine-tuned Turkish TurnGPT model, a zero-shot Turkish language model and a deployed multilingual detector are compared, together with voice-activity timing and facial features, under recording-disjoint and speaker-disjoint evaluation with paired cluster bootstrap tests. On held-out recordings, the fine-tuned model separates turn ends from pauses better than both text baselines and is the best calibrated of the three. Timing features of the current speaker are similarly informative, and their combination with the fine-tuned text score gives the best timing-based results; with logistic regression, this combination also outperforms the same combination with either text baseline. In a latency simulation, adding the text score to a silence timeout shortens the mean response delay at the same interruption rate. On automatic Whisper transcripts, the fine-tuned model retains its advantage over the zero-shot model.

cs.AI · cs.CL

版本更新 · v2arXiv:2610.07593

TRANSIT: Transparent Scale-in for Multi-Node LLM Training

Hyungyo Kim, Nicholas Satchanov, Hrishi Shah, Gaohan Ye, Jiaqi Lou, Robert Walkup, Shweta Salaria, I-Hsin Chung 等 12 位作者

原文摘要

The growing memory requirements of large language models (LLMs) increasingly force distributed training jobs to span multiple GPU nodes to accommodate their training state. This reduces per-GPU computational efficiency, increases network traffic, and exacerbates job queueing delays. To address these inefficiencies, we introduce TRANSIT, a system that transparently enables efficient distributed LLM training on fewer GPUs through memory oversubscription. TRANSIT optimizes CPU-GPU data movement during oversubscription by coordinating driver-managed page residency with execution-guided zero-copy access, exploiting recurring kernel execution patterns across training iterations. We evaluate TRANSIT on dense and mixture-of-experts (MoE) models across different parallelization strategies using up to 64 NVIDIA H100-80GB GPUs interconnected over RDMA over Converged Ethernet (RoCE). Compared with the minimum GPU count required to accommodate the training state in GPU memory, TRANSIT enables training on up to 50% fewer GPUs while retaining over 90% of the baseline per-GPU computational throughput. In some configurations, training on 17-33% fewer GPUs even improves throughput due to reduced distributed communication overhead. When training on fewer GPUs, TRANSIT achieves up to 243% and 87% higher throughput than conventional driver-runtime-managed memory oversubscription and framework-managed CPU offloading, respectively, while reducing CPU memory consumption by up to 79% compared with the latter. Training on fewer GPUs with TRANSIT also reduces per-node network traffic by up to 33% and improves throughput by up to 44% under constrained network bandwidth. Production-trace-driven simulations further show 15-49% lower inter-rack network traffic and up to 2-95% shorter job queue times for different cluster capacities.

cs.DC · cs.AR

版本更新 · v2arXiv:2608.07981

Distilling Physical Priors into Streaming World Models

Liangliang Zhao, Junying Wang, Danni Yang, Yifan Chang, Bin Fu, Yu Qiao, Bowen Zhou, Yihao Liu

原文摘要

Streaming world models predict future visual states online while maintaining physically coherent dynamics over long horizons. However, their rollouts often violate basic physical constraints. A common approach distills pretrained bidirectional DiTs into few-step causal generators. However, this paradigm suffers from two fundamental limitations: generic bidirectional teachers acquire limited physical priors from visually oriented pretraining, and the limited priors suffer further loss during bidirectional-to-causal distillation. We present PhyS, a three-stage framework for distilling physical priors into streaming world models. To acquire physical priors from real-world interactions, we construct PhyS-120K, a dataset of 120K real-world physical-interaction videos spanning rigid-body dynamics, soft-body deformation, fluid phenomena, and phase transitions. Each video is annotated with structured descriptions of object properties and causal state transitions. Physics-aware supervised fine-tuning injects the physical priors into a bidirectional 14B DiT teacher, which we then distill into a lightweight 1.3B causal DiT for few-step autoregressive streaming generation. Finally, we use online reinforcement learning to incentivize the distilled model to generate physically plausible rollouts and further propose Temporal Credit Routing (TCR) to address temporal credit assignment. TCR evaluates physical consistency over overlapping temporal windows and routes the resulting group-relative advantages to temporally aligned denoising actions. On PhysicsIQ, PhyS improves the Wan2.1-14B teacher by 18.2\% and the Self Forcing, Rolling Forcing, and Causal Forcing by 23.7\%, 14.8\%, and 31.4\%, respectively. Results also improve the physics-aware video benchmarks VideoPhy, VideoPhy2, and PhyGenBench. The dataset, code, and more sample videos are available on our Project Page.

cs.CV

版本更新 · v2arXiv:2611.16413

RICapBench: A Fine-Grained Visual Understanding Benchmark Dataset for Vision-Language Models in Remote Sensing

Xue Wang, Yubin Xu, Kun Tan

原文摘要

Advances in vision-language models (VLMs) have improved remote sensing image analysis and supported geospatial decision-making. However, fine-grained understanding of remote sensing imagery remains challenging for VLMs. Large variations in object scale, complex geospatial relations among instances, and limited visual detail hinder precise instance localization, visible attribute identification, and spatial relation understanding, potentially leading to descriptions that lack visual grounding. Existing datasets and benchmarks insufficiently assess end-to-end instance localization and captioning without target-specific textual cues. To support this evaluation, we introduce RICapBench, a dataset and benchmark that unifies instance localization and fine-grained captioning in remote sensing. RICapBench contains 245,466 instance-level annotations and 498,005 spatial relation tuples spanning general remote sensing, UAV remote sensing, and dense small-object scenarios. Each annotation pairs an object's oriented bounding box with a natural language description of its visible attributes and spatial relations to surrounding instances and environmental elements. To evaluate the accuracy and completeness of spatial relation captioning, we propose a remote sensing spatial relation parser and the RelCap-F1 metric. We combine RelCap-F1 with instance localization and attribute captioning metrics into a composite metric, RICScore. Our evaluation of 15 advanced open-source and closed-source VLMs identifies instance localization as a major bottleneck for these models. LocateAnything achieves the highest RICScore of 31.68, but scores lower on spatial relation captioning than other task-fine-tuned open-source VLMs. Code and dataset are available at https://github.com/Xldln/RICapBench.

cs.CV