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

+758新增论文
+578其中 AI 相关
+94已有论文修订
+95新增 AI 导读

arXiv 最近完成扫描:2026-11-17 04:58 · 全站最近入库:2026-11-17 04:58各信源进度 →

统计口径与数据新鲜度

统计 2026-11-16 10:07 — 2026-11-17 10:07(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:2026-11-17 04:58最近发现新增版本:2026-11-17 04:58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。

24,106 已收录论文20,158 AI 相关7,049 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 24,106 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-17 04:57(北京时间)
新论文 · v1arXiv:2611.09232

RAG-Enhanced LLMs for Interactive Explainability in Clinical Decision Support Systems

Lorenzo Mannocci, Francesca Naretto, Lucia Passaro, Anna Monreale

原文摘要

Artificial Intelligence (AI) is increasingly embedded in high-risks decision-making systems across domains such as healthcare, finance, and the judicial system. As reliance on predictive models grows, so does the demand for Explainable AI (XAI) to ensure transparency, trust, and interpretability. However, existing XAI approaches often fail to present explanations in a format accessible to domain experts, limiting their practical utility. Large Language Models (LLMs) have recently emerged as a bridge between complex models and end-users, offering natural language explanations. Yet, their use introduces new risks, including hallucinated outputs and lack of source traceability. To address these challenges, we propose a Clinical Decision Support System (CDSS) that not only combines standard predictive models with XAI techniques but also incorporates a Retrieval-Augmented Generation module as an interactive enhancer. This module grounds explanations in verified medical knowledge, reformulates technical outputs into a human-like format, and enables users-such as clinicians-to actively query the system. By fostering an interactive, human-in-the-loop environment, our approach empowers domain experts to explore model decisions, contextualize explanations, and build trust in AI-assisted diagnostics. The study concludes with preliminary experiments validating the proposed methodology.

cs.AI · cs.HC · cs.LG

版本更新 · v2arXiv:2607.05319

Steering Training Trajectories in Diffusion Representation Learning

Rajat Rasal, Avinash Kori, Tian Xia, Ben Glocker

原文摘要

Across training runs of a diffusion autoencoder, we observe similar image reconstruction quality, but substantially different levels of disentanglement in the learned representations. In this work, we study how training dynamics contribute to this behaviour. By analysing training trajectories in a controlled setting - curves of reconstruction quality against disentanglement during training - we identify two recurring trajectory patterns, which we call learning regimes. These regimes differ in how reconstruction and disentanglement evolve, despite similar reconstruction at the end of training. In the reconstruction regime, models improve image fidelity before disentanglement, while in the disentanglement regime, models improve both reconstruction and disentanglement more gradually. We hypothesise that modulating shortcut pathways in the diffusion U-Net and controlling early noise-level exposure can shape trajectories. To that end, we introduce SteeringDRL, combining gated residual U-Nets with a noise curriculum to steer models towards better disentanglement. Across attribute disentanglement benchmarks, SteeringDRL has competitive disentanglement, with reduced seed sensitivity in many settings. Our method extends to spatial disentanglement with object-centric learning, improving unsupervised segmentation on synthetic and real-world datasets.

cs.CV · cs.AI

版本更新 · v2arXiv:2610.40748

WorldBridge: Bridging Video and Action for Task Generalization in World Action Models

Jun Cen, Yubo Guo, Hao Li, Liangkai Li, Yaoxu Lyu, Meng Wei, Zeqi Xiao, Sizhe Yang 等 12 位作者

原文摘要

Achieving the broad task generalization of embodied models in physical environments remains a central challenge in robot learning. We study this challenge in World Action Models (WAMs) through two complementary capabilities: video generalization, which predicts the visual evolution of new tasks, and action generalization, which translates these predictions into executable behavior through a video-to-action bridge. By varying video and action supervision for tasks outside an initial training set, we examine their respective contributions to task transfer. On RoboTwin 2.0, using only in-distribution video-action supervision yields a 25.07% success rate on out-of-distribution tasks. Introducing videos of these tasks without their action labels raises success to 55.90%, while adding just one action-labeled demonstration per task further improves it to 70.70%. These findings suggest that broader video coverage enables task transfer by leveraging existing inverse-dynamics capabilities, while even sparse action supervision further strengthens those capabilities. Guided by this perspective, we introduce WorldBridge, a WAM with separate video and action experts connected through structured attention and trained with joint flow matching. WorldBridge incorporates visual memory at two temporal scales and action forcing, alongside a unified data processing pipeline that standardizes and filters heterogeneous robot trajectories and egocentric hand-derived supervision. Across multiple manipulation benchmarks, WorldBridge achieves state-of-the-art performance among the compared methods, including 96.23% on RoboTwin 2.0 full data setting and 80.9% on clean to random setting, 82.8% on RoboCasa GR1 TableTop, and 57.6% on DOMINO Level-1. Our study highlights the complementary roles of video prediction and action grounding in developing robots that generalize across tasks.

cs.RO

新论文 · v1arXiv:2611.09228

Rhythmic Fitness Beat 'Em All using Real-Time Monocular Human Pose Estimation

Yanis Zaoui, Valentin Bégel, Mohamed Adjel

原文摘要

Short and frequent physical activity exercices mitigate the health risks of sedentary behavior. Yet, the lack of motivation remains a critical barrier to adoption. Exergames gamify physical activity, but most rely on dedicated hardware, and outside dance and rhythm games they rarely use the musical beat. We present a work-in-progress beat 'em up exergame played with the whole body in front of a single smartphone, which runs the entire pipeline: an off-the-shelf real-time 3D human mesh recovery model on the phone's neural processing unit, gesture recognition, and the game, shown on any TV through a USB-C to HDMI cable. Our contributions are (1) a 3D recognition of five shadow-boxing moves (guard, punch, slip, squat, jump), each defined by one metric measure in a body frame aligned with gravity and one threshold with hysteresis, personalised during a tutorial; (2) a rhythm stage whose discrete cues are placed on the song's kick drum and judged on the movement's capture time, with a delay learned for each player; and (3) a phone-only architecture, running at 30 frames per second, that turns a phone and an HDMI cable into an exergame console.

cs.HC

新论文 · v1arXiv:2611.09225

Fused-Softmax for efficient quantized attention

Tianxing Wu, Juuso Korhonen, Jesus Carabano Bravo

原文摘要

Low-precision matrix throughput on modern GPUs is growing far faster than vector-arithmetic throughput. Attention therefore cannot realize the available GEMM performance because softmax has become the pipeline bottleneck. We introduce Fused-Softmax Attention, a training-free algorithm that reformulates and approximates softmax to fold its dominant operations into the attention GEMMs, moving work from vector units to matrix cores. Implemented for the AMD CDNA 4 architecture, our MXFP8, MXFP6 and MXFP4 kernels reach approximately 3.0, 4.0 and 4.6 PFLOP/s on an AMD Instinct MI355X GPU, yielding speedups of $2.1\times$, $2.8\times$ and $3.2\times$ over the BF16 attention. Importantly, the additional error introduced by Fused-Softmax remains below the error floor set by quantizing the GEMM operands. On Wan2.2-T2V-A14B, a hybrid MXFP6/4 schedule delivers up to $2.05\times$ end-to-end speedup while preserving the visual quality of the outputs. The kernels ship in AMD's open-source AITER library, and they powered AMD's MLPerf Inference v6.1 Wan2.2 submission, which ran $1.7\times$ faster than AMD's previous-round result on the same hardware. Our results show that redesigning softmax to utilize the matrix pipeline enables attention to capitalize on continued growth in low-precision matrix throughput.

cs.AI · cs.LG · cs.PF

新论文 · v1arXiv:2611.09224

HMMF-Net: A Lightweight Hierarchical Multi-Frame Multimodal Fusion Network for Accurate Fish Feeding Intensity Assessment

Phi-Hung Hoang, Van-Manh Tran, Nam-Thuan Trinh, Thi-Thu-Hong Phan

原文摘要

Automated fish feeding intensity assessment (FFIA) is important for precision aquaculture and feed cost optimization. However, unimodal methods are vulnerable to acoustic noise and visual disturbances, while existing multimodal frameworks often rely on computationally intensive 3D convolutions or Transformers, limiting their use on edge devices. To address these limitations, we propose HMMF-Net, a lightweight multimodal architecture that combines early multi-frame visual encoding and multi-axis acoustic modeling through a hierarchical bidirectional messenger cross-modal fusion (BMCF) mechanism. The visual branch captures temporal variations in feeding behavior by stacking temporally distant frames as a multi-channel input to an adapted 2D EfficientNetB0 backbone. The acoustic branch employs adaptive multi-axis residual units (AMARU) to extract temporal, frequency, and local time-frequency representations from log-mel spectrograms. BMCF progressively exchanges information between the two modalities at intermediate representation levels through compact messenger tokens, reducing the computational overhead of dense cross-modal attention. Evaluated on the AV-FFIA dataset using five-fold cross-validation, HMMF-Net achieves 96.77% accuracy and 96.88% F1-score on the held-out test folds, with 3.987M parameters and 1.002 GFLOPs. Compared with the best-performing visual configuration (95.49% accuracy), HMMF-Net reduces the classification error by 28.38%. Deployment on an NVIDIA Jetson Orin Nano using TensorRT with FP32 precision maintains comparable accuracy, with an inference time of 11.54 ms per 2-second sample and a model size of 18.81 MB. These results demonstrate that HMMF-Net combines effective multimodal representation learning with computational efficiency, enabling accurate and practical FFIA on resource-constrained edge platforms.

cs.CV · cs.AI

新论文 · v1arXiv:2611.09223

KLEAR: Knowledge-Matched LLM Evaluation in Korean across Application Requirements via Controlled Benchmark Transformation

Seoyeon Kim, Jeonghyun Kang, Hamin Koo, Barom Kang, Heondeuk Lee, Jaehyung Kim

原文摘要

Existing Korean benchmarks for Large Language Models (LLMs) evaluate performance across diverse knowledge domains, but their accuracy scores cannot distinguish whether incorrect responses arise from a lack of required knowledge or a failure to apply that knowledge correctly. Our analysis of existing Korean benchmarks reveals that most questions can be answered through factual recall alone, while those demanding greater application also differ in required knowledge, making it difficult to isolate application ability. To address this, we introduce KLEAR, a framework that constructs knowledge-matched contrasts by varying application demand while preserving the required knowledge. During construction, KLEAR identifies the knowledge underlying each source question as an explicit anchor bundle to guide generation of questions requiring more complex application. Generated candidates then undergo automated and human validation, with iterative revision based on evaluation feedback. Experiments demonstrate that the resulting knowledge-matched sets largely satisfy the intended constraints, enabling controlled evaluation across diverse LLMs. These evaluations reveal that increasing application demand exposes failures that persist even when the required knowledge is provided.

cs.CL

版本更新 · v2arXiv:2610.23287

I2V-DETACH: Source Grounding Detachment for Unauthorized Image-to-Video Generation

Chanhui Lee, Yeonghwan Song, Yewon Kang, Jeany Son

原文摘要

Recent diffusion models have advanced image-conditioned video generation, but their accessibility raises risks of unauthorized manipulation. Image immunization has emerged as a promising defense for Image-to-Video (I2V) protection, yet existing methods often rely on degrading visual quality or temporal coherence. In this work, we revisit I2V immunization as reducing source-content preservation, shifting the focus from output degradation to weakening source-conditioned content propagation. To this end, we propose I2V-DETACH, an image immunization method that prevents unauthorized I2V generation through source grounding detachment. I2V models preserve source content by propagating source-image cues into video latents during denoising. Accordingly, I2V-DETACH disrupts this process in two complementary ways: it suppresses source-video grounding to reduce source cue propagation, and strengthens intra-video interactions to shift denoising toward video-side dynamics with reduced reliance on the source image. To obtain informative optimization signals, we initialize auxiliary video latents with source-oriented states that remain within the source-conditioned generation regime while deviating from source-preserving trajectories. Extensive experiments across diverse diffusion architectures show that I2V-DETACH substantially reduces source-content preservation, improves over prior immunization methods, and exhibits strong black-box transferability.

cs.CV

新论文 · v1arXiv:2611.09222

Khmer PFGen-Bench: Building a Judge-Free Benchmark for Open-Ended Generation in a Low-Resource Language

Mengsay Loem, Kentaro Imajo, Ryota Ozaki, Masanori Hirano

原文摘要

Benchmarks for open-ended generation often rely on large language model (LLM) judges, whose reliability in low-resource languages is uncertain. We introduce Khmer PFGen-Bench, to our knowledge the first judge-free benchmark for multi-sentence answers to questions composed directly in Khmer, a low-resource language. Khmer PFGen-Bench adapts the scorer of PFGen-Bench, a benchmark for Japanese that scores responses with keyword rules and n-gram statistics from LLM-generated references. This scoring requires references that are both valid and varied, which LLMs may not produce in Khmer, so we examine how to build such references. We find that post-trained models can give more valid but less varied references than their base variants, and that mixing model families increases variety for the same number of references. Using references from post-trained models in nine families, the benchmark yields model rankings that are stable across reference sets and consistent with a native speaker's ratings. These findings offer guidance for building judge-free benchmarks in other low-resource languages.

cs.CL

新论文 · v1arXiv:2611.09217

Provable State Estimation with Recurrent Models

Vincent Andrieu, Pauline Bernard, Lucas Brivadis, Laurent Praly, Christian Wolf

原文摘要

Estimating the state of a dynamical system from observations is a crucial task in robotics and control. Despite their empirical success, recurrent models often lack formal guarantees on convergence and robustness. We bridge this gap by analyzing these architectures from the perspective of nonlinear observer theory. By framing RNNs as a time-discretization of continuous latent dynamical systems (Neural ODEs), we demonstrate that enforcing a contraction condition on the latent flow guarantees the existence of a smooth, injective map of the true physical state into the latent space. We prove that this architecture acts as a tunable observer: by increasing a single scalar gain that scales the latent vector field, the estimation error can be driven to an arbitrarily small neighborhood of zero at an exponentially fast rate. We further establish quantitative bounds on the MLP complexity required to decode this latent state, and extend our formal guarantees to the discrete-time residual RNN implementation. We validate the theory on a real mobile robotic task, where a provably contracting model outperforms a vanilla GRU with substantially more stable training. We also report that unconstrained GRUs converge to contracting dynamics on this task, suggesting that contraction is an emergent property of well-trained sequence models.

cs.LG · math.OC

新论文 · v1arXiv:2611.09214

Replication Certification at Publication: Checkable by Anyone, Anywhere, in Seconds, Forever

Gary King, Yiqing Xu, Leo Yang Yang

原文摘要

Authors increasingly deposit replication datasets---the data and code behind publications---in archives, and journals increasingly require them to, under progressively onerous rules. But a deposit alone proves little, so authors keep paying: months for the journal's reanalyses, years answering readers' questions, and rancor as failed replications become accusations because no record establishes what actually ran. We introduce a replication certificate: a cryptographically signed record of the deposited files and logged output. It replaces the journal's reanalyses with evidence anyone can check in seconds, which no one, including the archive, can backdate, and no one can fake without risking detection. Certification asks little of anyone and benefits everyone: The author deposits, then reruns the analysis once under a random seed the archive computes from the deposit's hash and a blockchain block; the archive signs, timestamps (in that blockchain), and stores the record---confidential data included---without rerunning anything; the journal merely checks a website for the certificate; and anyone can check or replay any step decades later.

cs.DL · cs.CR · stat.ME

版本更新 · v4arXiv:2610.31719

Two-colour crossing minimization in complete geometric graphs is NP-hard

Zhe Zhang, Yi Liu, Dan Yang

原文摘要

Given a fixed straight-line drawing of a complete graph, we seek a two-colouring of its edges that minimizes monochromatic crossing pairs. We prove that the threshold decision problem is NP-complete, even for points in general position with integer coordinates. Thus exact minimization is NP-hard, resolving the complexity question posed at the Crossing Number Workshop in 2019. We reduce from MaxCut on permutation graphs by representing source vertices with classes of geometric edges, whose crossings encode source adjacencies. The target complete graph also contains every other edge between its points, and their freely chosen colours can affect the encoded cost. To control these interactions, we construct a large comparison grid with two stable colour phases and use program points to turn this reference into colour preferences on other edges. The optimum differs from a computable baseline plus the amplified minimum number of uncut source edges by an error small enough to distinguish consecutive source values. The reduction is deterministic, with polynomially many points and polynomial coordinate bit length.

cs.CC · cs.CG

版本更新 · v2arXiv:2512.04992

Evolutionary Architecture Search through Grammar-Based Sequence Alignment

Adri Gómez Martín, Felix Möller, Steven McDonagh, Monica Abella, Manuel Desco, Elliot J. Crowley, Aaron Klein, Linus Ericsson

原文摘要

Neural architecture search (NAS) in expressive search spaces is a computationally hard problem, but it also holds the potential to automatically discover completely novel and performant architectures. To achieve this we need effective search algorithms that can identify powerful components and reuse them in new candidate architectures. In this paper, we introduce two adapted variants of the Smith-Waterman algorithm for local sequence alignment and use them to compute the edit distance in a grammar-based evolutionary architecture search. These algorithms enable us to efficiently calculate a distance metric for neural architectures and to generate a set of hybrid offspring from two parent models. This facilitates the deployment of crossover-based search heuristics, allows us to perform a thorough analysis on the architectural loss landscape, and track population diversity during search. We highlight how our method vastly improves computational complexity over previous work and enables us to efficiently compute shortest paths between architectures. When instantiating the crossover in evolutionary searches, we achieve competitive results, outperforming competing methods. Future work can build upon this new tool, discovering novel components that can be used more broadly across neural architecture design, and broadening its applications beyond NAS.

cs.NE · cs.AI · cs.LG

新论文 · v1arXiv:2611.09212

On the Limits of Execution Traces for Repairing Assertion Failures in LLM-Generated Code

Muntasir Adnan, Carlos C. N. Kuhn

原文摘要

Iterative repair returns a failed program's error to a large language model (LLM) and asks it to fix it. After iterative repair, the remaining failures are mostly assertion failures: with four Qwen2.5 models, they make up 79% of residual failures on BigCodeBench, all on HumanEval and 93% on LiveCodeBench. We argue, through an information view of the feedback, that an assertion failure says little about where the fault lies. We propose Breakpoint Debugging (BPD), which returns the failed program annotated inline with its variables' runtime values. We take the tasks that iterative repair failed to fix with five repair attempts and compare two ways of spending five further generations on them: BPD, and resampling. On BigCodeBench, BPD repairs 26.7% of these tasks against 22.4%, and its advantage grows with model size. On HumanEval, BPD outperforms resampling but only slightly, and on LiveCodeBench, fresh sampling repairs more (6.5% against 4.9%). Furthermore, BPD reduces unchanged-code resubmissions from 28% to 8% on BigCodeBench. Using only the first failing test's trace, it outperforms tracing all failing tests while using fewer tokens than both all-trace BPD and plain iterative repair. On LiveCodeBench, BPD gives no gain even on the public tests, so a visible failing input is not enough for a trace to help. These results match the information view: a trace helps only when the test exposes a failing input, and the fault is an implementation slip; otherwise, a repair loop should resample rather than trace.

cs.SE · cs.AI

版本更新 · v2arXiv:2611.01617

Seeing What Matters, Verifying What Pays: Load-Adaptive Speculative Decoding for Vision-Language Models

Hong Nie, Zhangchi Gao, Yuqian Liang, Jun Yu, Bo Wang, Jianqing Sun, Shengping Liu

原文摘要

Speculative decoding accelerates vision-language model (VLM) inference without altering the output distribution, yet two factors limit its gains. Existing drafters either lack the image evidence required for visually grounded tokens or pay to condition on every visual token, although a short draft block typically depends on only a few image regions. Moreover, the value of verifying additional candidates varies with GPU load, so no fixed verification structure is efficient across serving conditions. We present Uni-LaSpec, a block-parallel speculative decoding framework that addresses both factors. To improve draft quality, Uni-LaSpec reuses the verifier's own attention: the query retained from the preceding verification ranks the cached, int8-quantized visual keys, and a cross-layer vote selects a few visual tokens that the drafter fuses through gated cross-attention. Acceptance-aware training further weights each draft position by its contribution to the accepted prefix. To allocate verification, a marginal-cost criterion selects the structure per request and per round: draft trees exploit idle verifier capacity under light load, whereas a dynamic head truncates linear drafts under heavy load. Across six VLMs from three model families and 17 text and multimodal benchmarks, selecting eight visual tokens retains at least 98.6% of the acceptance obtained from the full image at sub-millisecond selection cost. With greedy decoding at concurrency one, Uni-LaSpec achieves up to 5.11x text and 3.24x multimodal throughput over autoregressive decoding. Across six load levels and three GPU types, it outperforms autoregressive decoding in every setting and matches or exceeds the best fixed verification strategy in 123 of 144 settings; under bursty arrivals, it reduces mean end-to-end latency by 47.6% relative to linear verification.

cs.AI · cs.CV

版本更新 · v3arXiv:2512.00791

Limitations of Using Identical Distributions for Training and Testing When Learning Boolean Functions

Jordi Pérez-Guijarro

原文摘要

When the distributions of the training and test data do not coincide, the problem of understanding generalization becomes considerably more complex, prompting a variety of questions. Prior work has shown that, for some fixed learning methods, there are scenarios where training on a distribution different from the test distribution improves generalization. However, these results do not account for the possibility of choosing, for each training distribution, the optimal learning algorithm, leaving open whether the observed benefits stem from the mismatch itself or from suboptimality of the learner. In this work, we address this question. That is, we study whether it is always optimal for the training distribution to be identical to the test distribution when the learner is allowed to be optimally adapted to the training distribution. Surprisingly, when we restrict our attention to efficient learners, i.e., those running in polynomial time, and assume the existence of one-way functions, the answer is negative. In other words, matching the training and test distributions is not always optimal.

cs.LG · cs.AI

新论文 · v1arXiv:2611.09210

TPS-A*: Path Planning for Any-Shape Vehicles with Certified Collision Checking in Trajectory Parameter Space

Jose Luis Blanco-Claraco, Shravan S. Rai, Fernando Cañadas-Aránega, José Luis Torres-Moreno, José Carlos Moreno-Úbeda

原文摘要

Global planners for ground vehicles usually approximate the vehicle by a disk, trading safety against conservativeness, while exact-footprint checking pays for every candidate motion, pose, and footprint edge. We present TPS-A*, a lattice planner whose edges are kinematically feasible trajectories from parameterized trajectory generators, built on certified collision checking in trajectory parameter space. A grid precomputed per footprint returns, in one pass over the local obstacles, a certified free distance along every candidate trajectory, so the per-node collision cost is independent of footprint complexity and trajectory count. We prove that this grid never overestimates the free distance of the continuously swept, possibly non-convex footprint, and state the planner's guarantees: collision-free paths and bounded suboptimality relative to its lattice abstraction. Goals with or without heading are reached exactly by Reeds-Shepp shots certified as trajectories of the same generators. An exact independent test finds no collision in its paths. On 300 cluttered worlds, with the true footprint and the same goal pose for all planners, it solves every query with a collision-free goal pose in a median of three milliseconds on an AMD Ryzen 7 2700X, on par with hybrid A*, 45\% of whose paths overlap obstacles, and nearly sixty times faster than a state-lattice planner, the only other collision-free one; with position goals, eighty-nine times faster than hybrid A*. It matches hybrid A* in automated parking and on narrow doorways, docks a non-convex fork footprint that no convex model admits, and drove a real robot across a university campus among people.

cs.RO

版本更新 · v2arXiv:2610.25863

On Multi-label radio Labelings of Graphs

Olivier Togni

原文摘要

We propose and study a multi-labeling generalization of the $k$-radio labeling problem of graphs: given a graph $G$ and integers $k$ and $m$, assign to each vertex of $G$ a set of $m$ integer labels such that distinct labels assigned to a vertex differ by at least $k+1$, and the labels of any two distinct vertices $x$ and $y$ pairwise differ by at least $k+1-d(x,y)$, with $d(x,y)$ being the distance between vertices $x$ and $y$. We prove general bounds on the minimum span of such a labeling and more refined results for the case $k=1$. For paths and cycles, we derive sharp bounds on the minimum span when $k$ is small. We finally propose a conjecture for the minimum span on the path and for an almost-tight bound on the cycle.

math.CO · cs.DM

新论文 · v1arXiv:2611.09207

RPEG: Training Reasoning Language Models Scalably and Cost-Effectively

Maciej Besta, Leonard Schmidt, Pierre Pang, Patrik Okanovic, Robert Gerstenberger, Torsten Hoefler

原文摘要

Reinforcement learning (RL) is now central to training Reasoning Language Models (RLMs), but RLM pipelines are difficult to scale and costly to run because each RL step couples multiple large models, heterogeneous stages, cross-step state, and complex asynchronous dependencies. Existing systems typically optimize only parts of this design space. We address this complexity by jointly optimizing the pipeline structure as well as its execution using a novel data-centric graph intermediate representation called RLM Pipeline Execution Graph (RPEG). RPEG exposes structural choices such as bounded asynchrony and actor-critic sharing together with execution optimizations such as stage fusion, fine-grained scheduling, and multi-dimensional model parallelism. It is coupled with a comprehensive analytical performance model for RLM training, jointly capturing compute, communication, memory, and hierarchical rack-, node-, and GPU-level effects. Across systems with up to 256 GH200 GPUs, our distributed runtime engine implementing RPEG improves performance by up to 2x over ReaL and reduces step latency by up to 2.48x over verl, while increasing hardware utilization by up to 30%. Overall, RPEG provides a graph-driven foundation for next-generation training pipelines for frontier reasoning LLMs.

cs.DC · cs.AI · cs.CL · cs.LG · cs.PF

版本更新 · v3arXiv:2608.22788

TailSieve: Partial-Rollout-Guided Tail Routing for LLM Rollouts

Tianqi Xu, Lu Lv, Haoyang Huang, Wenjie Huang, Zhanming Shen, Yuhao Shen, Baolin Zhang, Xinyi Hu 等 18 位作者

原文摘要

Large-scale rollouts have become a core component of modern LLM systems, spanning reinforcement learning (RL) post-training, on-policy distillation, and sampling-heavy evaluation. Unlike online serving, which typically targets request-level latency and throughput, synchronous rollouts must collect a complete batch before proceeding. A few long-tail generations can therefore dominate an entire rollout step, and uniform routing can slow these generations further by placing them in high-concurrency decoding batches. We present TailSieve, a partial-rollout-guided framework that jointly controls tail routing and replica allocation. Motivated by the observed persistence of prompt-level tail tendencies across policy updates, TailSieve uses partial-rollout completion status as a training-free signal for identifying candidate tail groups. Selected prompts are combined with newly sampled prompts, and all responses are generated from scratch under the current policy. A hierarchical controller uses response-work history and a serving model: a fast inner loop adjusts the number of isolated groups to balance completion times, while a slower outer loop adapts how many replicas serve these groups. The resulting low-concurrency execution also enables route-specialized speculative decoding with MTP or DFlash. Across the evaluated configurations, TailSieve achieves up to $1.67\times$ speedup with routing alone and $2.59\times$ with routing and speculation combined, both over uniform group routing.

cs.AI · cs.LG

新论文 · v1arXiv:2611.10067

The Challenge of Modeling the Complexity of Use for the Measurement of Digital Maturity in Education

Christine Michel, La{ë}titia Pierrot, Frederic Oru, Olivier Vigneau

原文摘要

Measurement of teachers' digital maturity can be useful in diagnosing technology adoption in education and in supporting teachers' professional development. However, there is a lack of easily usable tools in the existing literature. By investigating the most effective ways to measure teachers' digital maturity, our research aims to fill this gap. The focus of this article is on a data-driven approach to the assessment of digital maturity in education. Following a process similar to teaching analytics (data collection, data modeling, data understanding and visualization, and contextual understanding), we draw on diary data collected from 12949 French primary school teachers using a virtual learning environment (VLE) during the 2022--2023 school year. Specifically, the paper presents results related to modeling and visualizing data. To this end, we propose a critical analysis of two automatic classification techniques, one supervised rule-based and the other unsupervised (hierarchical clustering). In addition to providing different scales to measure, we have proposed a new approach to diagnose and visualize digital maturity. In this way, the method provides insight into the diversity and intensity of technology use, contributing to a better understanding of digital maturity and providing a practical tool for assessing day-to-day teaching practices.

cs.CY

新论文 · v1arXiv:2611.09205

Computation-Efficient Aerial-Marine Integrated Networks for Search and Rescue via Cooperative HAPS, UAVs, and MASSs

Zhen Wang, Bin Lin, Qiang Ye

原文摘要

In this paper, we propose an aerial-marine integrated network architecture tailored for maritime search and rescue (MSAR) operations, where a high altitude platform station (HAPS) and maritime autonomous surface ships (MASSs) are equipped with edge servers, providing computing services for surveillance uncrewed aerial vehicles (UAVs). These UAVs generate substantial volumes of computation-intensive and delay sensitive tasks, some of which can be strategically offloaded to both MASSs and HAPSs for distributed processing. We formulate the joint task offloading and resource allocation as a mixed-integer nonlinear program (MINLP) to minimize the system computation overhead (CO), quantified by a weighted sum of task completion time and energy consumption. Given the stochastic nature of task arrivals from UAVs and the dynamic characteristics of marine channel conditions, we employ a two-stage optimization to decompose the problem into four tractable subproblems, i.e., 1) edge server selection, 2) transmission power allocation, 3) edge computing resource allocation, and 4) local computation resource allocation. In stage I, we optimize the edge server selection via many-to-one matching. In stage II, we determine the transmission power of each UAV with quasi-convex optimization and solve the edge computing and local computing resource allocation subproblems via a projected gradient descent method and convex optimization, respectively. Simulation results validate that the proposed scheme can achieve superior system performance over benchmark schemes.

cs.NI

版本更新 · v3arXiv:2611.00800

A Discrete-Time Variational Objective for Degenerate Diffusion and Equivalence to Flow Matching

Rares Dolga, Harshad Dahake, David Barber

原文摘要

Diffusion models and more recently Flow models have widespread application. In this work we consider degenerate Diffusion models in which the per-step generator is deterministic. We construct a modified variational framework and show a formal equivalence between Flow models and Degenerate Diffusion models. For our discrete-time variational objective we show that the optimal degenerate generator recovers the data distribution in the limit of an infinite number of generator steps. Our work therefore provides a unified perspective, showing that discrete-time implementations of flow matching and the discrete-time Degenerate Diffusion model can be considered within the same simple variational framework, removing the need for the continuous-time machinery normally used to justify flow matching models.

cs.AI

新论文 · v1arXiv:2611.09204

Representational drift in a code model: behaviour settles, which neurons carry it does not

Piotr Wilam

原文摘要

In neuroscience, representational drift is a change in which neurons carry a representation while behaviour stays the same; we measure its analogue in OpenCoder-1.5B during pretraining, for 169 Python constructs over two trillion tokens at a constant learning rate, with a noise unit on every change, at checkpoints as close as 1,000 training steps. For 16 of the 169, those whose grammar fixes a next token, the loss at those positions reaches its floor within the first 4% of training, while the model's reliance on the neurons carrying them grows to the end: removing them costs 0.06, 0.25 and 0.40 nats more than removing a random set, at 4 billion, 84 billion and 2 trillion tokens. Which neurons carry a construct never settles: for all 169 the carrier set is still changing at the last checkpoint, and an abrupt settling at any date from 84 billion to 1.5 trillion tokens, or a gradual one before 500 billion, would have been detected in 99% of injected cases. The set measured at 84 billion tokens, removed at the end, cannot be separated from a random set, so the function is carried by whichever neurons are current. Up close the change is a persisting pool and a changing sample: a neuron that leaves a construct's set returns at three times the rate of a matched neuron, and no set reassembles. The hypothesis we fixed in advance, that the carriers settle, is refuted; we discuss candidate causes and name the experiment that would separate them.

cs.CL · cs.LG

新论文 · v1arXiv:2611.10066

VLM as an Editing Pilot: VLM-Steered Velocity Control for Flow-Based Image Editing

Kuan-Shen Chiu, Jia-Wei Liao, Yu-Chieh Chao, Li-Xuan Peng, Wei-Yun Ma, Cheng-Fu Chou

原文摘要

Text-guided image editing requires realizing the requested semantic change while preserving content unrelated to the instruction. FlowEdit provides an efficient inversion-free solution, but its source-anchored coupling can become overly restrictive, causing zero-edit, where the output remains close to the source and the requested change fails to emerge. We find that source coupling directly governs the preservation--editing trade-off, while a fixed coupling schedule cannot adapt to timestep-specific editing needs. Motivated by this observation, we propose V2C-Edit, a training-free VLM-steered velocity control framework. V2C-Edit constructs complementary coupling-controlled velocities and dynamically selects among them using Instruction-Level Similarity (ILS), which evaluates whether the predicted transformation follows the editing instruction. Adaptive spatial masking further preserves regions unrelated to the edit. Experiments on MagicBrush and PIE-Bench with FLUX and Stable Diffusion 3 show that V2C-Edit consistently improves instruction alignment over FlowEdit while maintaining competitive structural preservation, and that ILS better reflects editing effectiveness than target-caption CLIP similarity. Our code is available at https://github.com/Kuanshen77/V2C-Edit.

cs.CV

新论文 · v1arXiv:2611.09200

Structure and Lineage Evidence for Vietnamese Statutory Article Retrieval on DRiLL

Duc-Vu Nguyen, Ngan Luu-Thuy Nguyen

原文摘要

We study Vietnamese statutory Article retrieval on DRiLL using features for provision identity and local form, same-law support, and citation and lineage evidence. Across five random seeds, selectors sharing candidates, listwise Qwen evidence, and learner settings improve official $F_2$ from 71.77% to 72.91%, with the largest descriptive gain on two-Article questions. Added features account for 18.20% of the proposed selector's total decision-tree split gain, measured as training-objective reduction, mainly from same-law and citation evidence. Cases reveal base provisions and amendments, companion recovery, distractor removal, and failures. These findings motivate research on other datasets, representations, and learners without establishing individual feature effects or requiring LightGBM, hand-crafted scores, or explicit graphs.

cs.IR · cs.CL

版本更新 · v4arXiv:2501.13229

The Yoneda embedding in simplicial type theory

Daniel Gratzer, Jonathan Weinberger, Ulrik Buchholtz

原文摘要

Riehl and Shulman introduced simplicial type theory (STT), a variant of homotopy type theory which aimed to study not just homotopy theory, but its fusion with category theory: $(\infty,1)$-category theory. While notoriously technical, manipulating $\infty$-categories in simplicial type theory is often easier than working with ordinary categories, with the type theory handling infinite stacks of coherences in the background. We capitalize on recent work by Gratzer et al. defining the $(\infty,1)$-category of $\infty$-groupoids in STT to define presheaf categories within STT and systematically develop their theory. In particular, we construct the Yoneda embedding, prove the universal property of presheaf categories, refine the theory of adjunctions in STT, introduce the theory of Kan extensions, and prove Quillen's Theorem A.

cs.LO · math.AT · math.CT

新论文 · v1arXiv:2611.09199

SAS: Scene-Annotated AudioSet Strong for Joint Scene-Event Analysis

Michel Olvera, Paraskevas Stamatiadis, Romain Serizel, Slim Essid, Gaël Richard

原文摘要

Context is central to human auditory perception, as listeners interpret sounds within their acoustic environment to resolve ambiguity. However, large-scale datasets that jointly capture sound events and acoustic scenes remain limited. We address this by augmenting AudioSet Strong with clip-level scene annotations under two taxonomies for joint sceneevent modeling. A corpus-scale analysis shows that scene identity is better captured by events distinctive to a scene than by those that merely occur frequently. Reweighting event co-occurrences toward scene-specific associations produces representations that better align with human-derived taxonomies, while also exposing scenes where the acoustic and semantic views diverge. This geometric divergence predicts classification errors, while supervised validation confirms that the proposed scene annotations provide meaningful complementary information for developing scalable context-aware machine listening systems.

cs.SD

版本更新 · v2arXiv:2611.06984

AdaFloor: Adaptive Floor Planning and Rank-Time Reuse for Long-Tail MoE RL Rollouts

Jia Wei, Ziyi Qiu, Yi Xu, Xuehai Qian

原文摘要

Synchronous MoE reinforcement-learning post-training suffers from response-length imbalance. Existing methods reduce request tails, yet completed ranks remain bound to Expert Parallelism (EP) collectives and expert ownership. In a representative 16-accelerator Qwen3-30B-A3B rollout, this dummy run traps 32.83% of rank-time. AdaFloor converts this EP-coupled capacity into bounded leases for auxiliary work to improve system resource efficiency through rank-time reuse. Historical prompt-tail prediction and a calibrated guard construct the KV-safe layout. AdaFloor then selects the floor-specific safe quorum schedule with the largest predicted release area. Its runtime shrinks the EP group and leases detached ranks until full-world restoration. Across three updated-policy seeds on DeepScaleR mathematical reasoning rollouts, its capacity-oriented Natural floor2 configuration with TailGuard reduces Vanilla's 29393.9 rank-seconds of dummy run by 96.15%. Rollout shortening eliminates 18693.6 rank-seconds, while coordinated exits convert another 9568.6 rank-seconds into reusable capacity, leaving 1131.7 rank-seconds. For sidecar reuse, its Planned residency floor4 configuration without TailGuard serves 3134 Qwen3-8B GSM8K queries per epoch at 99.46% accuracy and incurs only a 2 to 3% primary throughput reduction against the same configuration with sidecar execution disabled. Its latency-oriented Planned residency floor4 configuration with TailGuard reduces rollout time by 24.09% relative to Vanilla and by 2.31% relative to LengthSort with TailGuard. On public Eurus-2-RL code rollouts with the same model and 16-rank topology, AdaFloor reduces rollout time by 20.56% and dummy run by 89.13% while generated work and mean reward remain stable.

cs.AI · cs.DC

版本更新 · v3arXiv:2609.08572

AgentGrad: Intervention-guided Prompt Optimization for Multi-Agent Systems

Jaewon Chu, Jinwoo Seo, Jaewon Cho, Jeehye Na, Yunyang Xiong, Youngdae Kim, Hyunwoo J. Kim

原文摘要

Large language model (LLM)-based multi-agent systems (MAS) achieve strong performance by employing specialized multiple agents, yet their performance depends on the prompt design of each agent. For MAS prompt optimization, textual gradient methods that guide prompt updates using natural-language feedback have emerged as a leading paradigm. In this paper, we identify limitations in two stages of existing textual gradient approaches: gradient extraction and gradient aggregation. In gradient extraction, previous works select a prompt to update without verifying whether modifying it resolves the failure, and derive gradients without agent-level supervision over the corresponding agent's intermediate output. In gradient aggregation, individual gradients are randomly grouped and concatenated, often mixing unrelated failure modes and producing prompts that fail to generalize. To address these limitations, we propose AgentGrad, a prompt optimization framework for multi-agent systems based on sequential intervention and semantic textual gradient abstraction. For each failure, sequential intervention injects a hint into one agent at a time to identify the target agent whose modification resolves the failure. The modified output of the target agent then serves as agent-level supervision for extracting a fine-grained gradient. Semantic textual gradient abstraction clusters semantically similar gradients to prevent mixing unrelated failure modes, and abstracts each cluster into a generalized gradient that captures the shared corrective pattern. Experimental results show that AgentGrad achieves state-of-the-art performance across five MAS benchmarks while reducing wall-clock optimization time by 2.5 times and optimization cost by 21.8% on average compared to the next-best baselines. Our code is available at https://github.com/mlvlab/AgentGrad.

cs.AI