跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+479新增论文
+361其中 AI 相关
+63已有论文修订
+93新增 AI 导读

arXiv 最近完成扫描:2026-11-16 01:34 · 全站最近入库:2026-11-16 13:44各信源进度 →

统计口径与数据新鲜度

统计 2026-11-15 16:09 — 2026-11-16 16:09(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-16 01:34最近发现新增版本:2026-11-16 13:44库内论文最新更新时间:2026-11-14 02:59

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

23,827 已收录论文19,941 AI 相关6,921 有修订112 HF 社区发现

浏览论文库

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

找到 23,827 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-16 01:33(北京时间)
版本更新 · v2arXiv:2611.11868

Fade to... Red? Scheduling Coloured Noise Exploration Supports Efficient Robot Learning with Reduced Movement Jerk

Francisco M. López, Francisco Cruz

原文摘要

Exploration in deep reinforcement learning (RL) for continuous control is commonly driven by temporally uncorrelated Gaussian (white) action noise, which leads to jerky motions with locally constrained exploration. In contrast, temporally correlated coloured noise, particularly pink noise, has been shown to support more efficient exploration in RL. Motivated by these findings, we inquire whether scheduling changes to the noise colour over the course of training can outperform white noise for continuous control robotics. We perform a systematic study of 49 linear schedules between white and red noise, together with generalised state-dependent exploration (gSDE), on four sparse-reward manipulation tasks with a Fetch arm and a Shadow Dexterous Hand. We find that the noise colour at the end of training does not affect learning efficiency or end performance, whereas fading towards red noise halves the jerk of exploration in every task. Balancing learning efficiency, end performance and smoothness, red noise emerges as the preferred strategy: it learns as efficiently as pink noise with almost 2-fold less movement jerk. We therefore recommend exploration schedulers that fade towards red noise as the default for off-policy robot learning.

cs.RO · cs.LG

新论文 · v1arXiv:2611.13621

PUMA: Probability-Constrained Sequential Interaction for Forensic Localization

Zhouli Yu, Mengyuan Zhou

原文摘要

Image manipulation localization must jointly exploit semantic structure and noise-sensitive forensic evidence. However, the effects of spatial correspondence, matching quality, and interaction order are often conflated. Existing semantic-forensic fusion methods provide complementary representations, but do not explicitly preserve quality-mismatch information during spatial correspondence or adaptively select the order in which information is propagated across branches. We propose PUMA (Probability-constrained Unbalanced Matching and Aggregation), which combines spatially constrained unbalanced optimal transport with position-wise convex selection between two sequential operators through a shared positional interface. Before normalization, PUMA retains signed transport-mass residuals, constructs row-stochastic cross-branch and within-noise-branch operators, and updates noise features through a gated residual path. PUMA achieves strong localization results on four test sets, with scores on COVERAGE and NIST16 exceeding the published results of the representative methods listed in our comparison. The experiments support the utility of local matching states and adaptive sequential interaction for cross-source localization and several post-processing conditions.

cs.CV

新论文 · v1arXiv:2611.13620

When to Start Denoising? Adaptive Denoising Entry for Generative Video Compression

Zifu Zhang, Yuqiao Li, Ding Ding, Shengxi Li, Mai Xu, Daowen Li, Kai Li, Ying Chen

原文摘要

Generative video compression offers a promising solution for compressing and reconstructing videos under extremely low bitrate constraints, among which the diffusion-based solutions still achieve the best performance by providing the powerful generative latent prior. However, their effectiveness critically depends on when the compressed latent is introduced into the diffusion process, which directly governs the trade-off between reconstruction fidelity and perceptual realism. To address this limitation, we identify denoising start-timestep selection as a key factor governing this trade-off and formulate it as a geometric optimization problem. Based on this formulation, we propose a novel adaptive diffusion entry method for generative video compression (ADE-GVC), the first timestep-aware generative video compression framework, featuring a lightweight predictor that estimates an adaptive denoising start point from the compressed latent. A temporal adapter further extracts compression-aware conditions and injects them into a pretrained video diffusion model, enabling the reconstruction of videos with both high fidelity and realism. More importantly, we observe a strong quantitative relationship between the predicted timesteps and the distortion between the original and compressed latents, providing empirical support for our optimization formulation. Extensive experiments demonstrate that ADE-GVC achieves the state-of-the-art performance on perceptual metrics and substantially improves reconstruction realism over existing generative video compression methods. Notably, our ADE-GVC achieves an average bitrate savings of 40.99\% (on DISTS) and 95.24\% (on PSNR) compared to the latest open-source baseline GLC-Video.

cs.CV

版本更新 · v6arXiv:2610.16717

Where Is the "Self" of a Large Language Model? -- From Self-Index to Self-Reference

Yuchen Li, Xintong Li, Zongqi Fan, Ken-tye Yong, Nguyen H. Tran

原文摘要

Whether a large language model has a self is usually argued without a way to measure it. We measure a functional self with two levels familiar from biology: a self-index, which marks material as the model's own, as animals tag self-generated signals with a corollary discharge, and self-reference, the ability to recognize what is one's own, as in mirror self-recognition. Our main finding is that language models mark their own messages internally and that recognizing them depends on this mark: no model we tested recognizes its own messages without it, and in the four models we edited, changing the mark removes or creates the recognition. With the same text placed in an Assistant, a User, or a System message of its chat template, Qwen3.8-27B separates its own message from both external ones at all 64 layers and in every held-out text. Asked who wrote a message, it answers "you" with probability 0.77 for its own messages and at most 0.07 for the others. Setting eight coordinates per layer, along which its own message differs most from the same text as a User message, to their User values lowers the rate of "you" to at most 0.05, and the reverse edit makes the model claim a User message, closing at least 41% of the gap. Random directions do neither, and answers about the content stay unchanged. This holds in Qwen3.8-27B without and with thinking, gpt-oss-20b, gpt-oss-120b, and MiMo-V2.6-Distill-Qwen-9B. Across twenty-five open-weight models from fourteen developers, the two levels are nested: no model recognizes its own messages without a self-index (permutation p = 0.004). Without any question, next-token prediction on the model's own message differs from that on the same text in another role, more so as more layers carry the self-index (Spearman rho = 0.82), and the same coordinates carry this difference. These results support the account that self-reference is a readout of a self-index.

cs.CL

版本更新 · v2arXiv:2605.30770

SSR: Scaling Surefooted and Symmetric Humanoid Traversal Toward the Open World

Ruiqi Yu, Yiwen Wang, Yuan Hao, Jun Wu, Qiuguo Zhu

原文摘要

Advancing humanoid traversal toward the open world is key to practical deployment in human environments, but remains challenging. The robot must use vision to ensure safe and reliable foot placement on heterogeneous terrain during highly dynamic motion, while producing coordinated, natural whole-body behaviors. We propose SSR, an efficient end-to-end framework for egocentric vision-based humanoid traversal that jointly learns these capabilities. SSR introduces imagined foothold guidance, which learns to model forthcoming swing-foot contacts and evaluates their support to guide pre-touchdown swings toward stable regions, reducing edge slips. It further employs equivariant latent-space symmetry augmentation to efficiently induce bilateral coordination under high-dimensional visual observations, and uses terrain-specific multi-discriminator motion priors to encourage human-like behavior across scenes. Extensive experiments show that SSR achieves safe, stable, and high-quality locomotion on diverse real-world terrains, including stairs with varied structures and extreme challenges such as wide gaps and high platforms, while enabling reliable long-horizon traversal in diverse outdoor environments. Project website: https://ssr-humanoid.github.io/.

cs.RO

版本更新 · v2arXiv:2611.11821

Direct Trajectory Optimization for Inference-Time Alignment of Text-to-Image Diffusion Models

Huixuan Zhang, Li Lin, Junzhe Zhang, Xiaojun Wan

原文摘要

Alignment of text-to-image diffusion models with reward functions is a key challenge in controllable generation. Existing approaches typically rely on optimizing model inputs or fine-tuning denoising networks, which requires computationally expensive backpropagation through large denoising models. In this work, we propose Direct Trajectory Optimization (DTO), a training-free framework that performs alignment by directly optimizing the sampling trajectory of diffusion models. Specifically, DTO treats intermediate denoising scores and sampled noise as the optimization variables, eliminating the need for gradients through the denoising network. To mitigate the out-of-distribution problem, we introduce Reward-Optimized Score Guidance, which interpolates between the optimized scores and the pretrained model predictions to stabilize sampling, while providing a provable theoretical guarantee of reward improvement under constrained distributions. Empirically, DTO improves alignment performance across three diffusion models on two benchmarks, outperforming existing inference-time alignment methods while maintaining high efficiency. These results suggest that trajectory-level optimization is a promising direction for efficient and effective diffusion model alignment.

cs.CV · cs.AI

新论文 · v1arXiv:2611.13615

Synapse-Control: Making Agent Experience Actionable with Lightweight Runtime Control

Zhiyuan Huang, Shengan Zheng, Chenghu Yang, Song Zhang, Lingyun Li, Chen Yu, Weiquan Huang, Linpeng Huang

原文摘要

Large language model agents reuse execution experience to improve reliability on complex tool-based tasks. Systems such as Sentry call large models to generate failure diagnoses before retrieving advice, adding inference cost to experience reuse. We present Synapse-Control, an agent runtime whose lightweight control plane selects advice and determines when to send it to the task-executing model. Rules shortlist relevant advice, and a lightweight 1.7B classifier fine-tuned on teacher-labeled state-suggestion pairs scores those candidates instead of generating a diagnosis. Micro-contracts encode reusable interface checks, such as required arguments or valid options, without prescribing complete task programs. A Prompt Value Table records earlier guidance and feedback to limit repeated suggestions through cooldown and token-budget checks. Across 1,308 attempts per method in AppWorld and WebShop, Synapse-Control achieves 270 successes (29.8% more than Base), replaces 4,987 generative diagnosis calls, and reduces auxiliary control time relative to Sentry by 50.2% and 51.0%, respectively.

cs.CL · cs.AI · cs.LG

版本更新 · v2arXiv:2607.28422

QAdapt: Noise-Adaptive Neural Pre-Decoding for Surface-Code Quantum Error Correction

Ran Miao, Rui Luo, Xiaohan Shan, Xiaoming Sun

原文摘要

Fault-tolerant quantum computing relies on quantum error correction to protect logical information, requiring classical decoders that keep pace with syndrome generation. Heterogeneous, changing hardware noise and the distribution shift from simulation to experiment challenge the accuracy and robustness of fixed neural decoders. We present QAdapt, a noise-adaptive neural pre-decoding framework for surface-code quantum error correction. QAdapt learns local spatiotemporal correlations, continually adapts across noise environments, and passes the residual syndrome to a conventional global decoder. Online elastic weight consolidation limits forgetting as training moves between measurement, idle, CNOT, and Pauli-bias noise regimes. At distance nine, the HTNet pre-decoder reduces equivalent per-round logical error by 15.64% relative to Ising-Fast, with 28.75% fewer parameters and 9.07% less matching-backend time. Online EWC reduces average forgetting by 76.52% relative to sequential fine-tuning. With weights frozen, it reduces equivalent per-round logical error by 5.66% relative to continued base-environment training on a fixed valid domain of 331 unseen compound-noise conditions; whole-sequence failure probability falls by 2.54% over all 396 conditions. On native Willow records, the corresponding per-round reductions are 19.44%, 17.47%, and 17.95% at 10, 30, and 50 rounds, without target-domain fine-tuning. These results show how adaptive neural pre-decoding can improve logical accuracy, preserve performance across noise regimes, and reduce residual matching workload within an established decoding pipeline.

cs.LG · quant-ph

新论文 · v1arXiv:2611.13614

Emilia2: A Large-Scale Dataset for Controllable Long-Form Speech Generation

Qinke Ni, Dekun Chen, Junwen Qiu, Huan Liao, Xu Li, Xingwen Han, Wan Lin, Kunyu Feng 等 17 位作者

原文摘要

Generating long-form speech requires models to keep content, speaker identity, turn-taking, and expressive style consistent over time. Yet most open training corpora contain examples lasting only tens of seconds and open long-form benchmarks cover at most two speakers per recording. We introduce Emilia2, a 229K-hour multilingual dataset for controllable long-form speech generation. From in-the-wild data, Emilia2 constructs short utterances, single-speaker long-form speech, and multi-speaker dialogues, preserving transcripts, speaker identities, pauses, and acoustic transitions. Examples longer than five minutes account for 50.0% of single-speaker long-form hours and 66.8% of dialogue hours. To support expressive control, Emilia2 provides natural language instructions for 4.77M examples and nonverbal-vocalization tags for 933K examples. We also introduce Emilia2-testset, 23 hours of bilingual recordings of up to 30 minutes and four speakers, each speaker paired with reference audio. We train autoregressive (AR) and non-autoregressive (NAR) models on Emilia2. Our 0.6B-backbone NAR model achieves the lowest Chinese CER and English WER on TTSD-Eval (5.27%/9.93%) among compared systems, while our AR model is on par with Qwen3-TTS-12Hz-VD on InstructTTSEval (zh/en: 76.2/78.3 vs. 77.1/77.9). These results validate Emilia2 as a comprehensive data foundation for long-form speech generation and minute-scale expressive control. Demo: https://charlesnii.github.io/Emilia2-demo/.

eess.AS · cs.SD

新论文 · v1arXiv:2611.13613

Megrez: A Knowledge-Guided Agent for Formal Mathematics over MK Set Theory

Tiancheng Xu, Wensheng Yu

原文摘要

General-purpose coding agents equipped with proof-assistant tools have shown strong performance on standalone competition problems, yet cumulative textbook formalization inside a low-resource mathematical library still faces three obstacles: aligning textbook concepts with library declarations, keeping long-horizon search synchronized with live proof states, and integrating locally verified proofs reliably into a growing codebase. We present Megrez, a knowledge-guided formalization agent for Rocq libraries built on Morse--Kelley (MK) set theory. Megrez coordinates a general coding agent through three mechanisms: (1) a source-rebuildable knowledge graph merging formal declarations, compiler references, and curated metadata to support retrieval via Chinese aliases, concept clusters, proof dependencies, and multi-lane full-text ranking; (2) a feedback-driven 1--5 tactic short-burst proving loop paired with a goal-local branch ledger and exponential-interval structured reflection; and (3) a three-check acceptance pipeline (statement fidelity, axiom auditing, and full builds) coupled with a multi-corpus Provider stack and a candidate-bound recoverable batch state machine. Deployment and evaluation on a point-set topology corpus demonstrate effective cross-chapter formalization and scalable premise retrieval.

cs.AI

新论文 · v1arXiv:2611.13612

Earth Embeddings in Environmental Remote Sensing

Yuchi Ma, Jiaqi Yang, Mengyu Liang, Ivan Higuera-Mendieta, Caleb Robinson, Anthony Ortiz, Min Chen, David B. Lobell 等 10 位作者

原文摘要

Earth observations from satellite platforms have profoundly transformed research and applications across numerous domains. However, uneven observation coverage and frequent data gaps continue to limit their ability to capture Earth's surface dynamics. Moreover, processing multi-modal Earth observations requires substantial computational resources and sophisticated data harmonization frameworks, which may not be readily accessible to many remote sensing practitioners. Earth embeddings have emerged as promising solutions to these challenges. As general-purpose representations converted from multi-source Earth observations, Earth embeddings can support a wide range of downstream tasks with no requirement of task-specific feature design and preprocessing. In this study, we first introduce the theoretical foundations of Earth embeddings. Then, we review several representative Earth embedding products. We next examine the applications of Earth embeddings across six major domains of environmental remote sensing. Finally, we systematically assess the performance of Earth embeddings in four downstream tasks, including hyperspectral image classification, individual tree crown segmentation, crop yield prediction, and aboveground biomass mapping. Our review and case studies suggest that, despite several advantages over conventional Earth observation data, current Earth embeddings do not always outperform remote sensing imagery or hand-crafted features. Earth embeddings should not be viewed as a direct replacement for conventional EO data, but rather as representations that provide stable information across space and time. Remote sensing practitioners may consider evaluating Earth embeddings' suitability for specific tasks and using them strategically, rather than assuming superior performance by default.

cs.AI

版本更新 · v2arXiv:2605.12798

Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment

Baris Askin, Muhammed Ustaomeroglu, Anupam Nayak, Gauri Joshi, Guannan Qu, Carlee Joe-Wong

原文摘要

Post-training a language model can induce surprising behavioral changes far beyond the intended training target. We study two particularly striking examples: emergent misalignment (EM), where fine-tuning on a narrow dataset leads to misaligned behavior in unrelated contexts, and subliminal learning (SL), where a student model inherits a teacher's misalignment through seemingly benign training data. We examine how the training data, evaluation data, and training method shape where and how strongly misalignment transfers between data and models. We characterize the data by the domain of a prompt (its topical context) and its task (the function the model performs), and we build natural-language and synthetic datasets organized as domain-task grids. We find that EM transfers more strongly to prompts that share the task of the fine-tuning data than to prompts that share its domain. In the synthetic setting, tasks that the model has learned well spread misalignment more broadly. For SL, we go beyond SFT (supervised fine-tuning) to off-policy distillation on the teacher's full next-token distribution and, to our knowledge for the first time, on-policy distillation, where the teacher only scores tokens sampled by the student. Misalignment transfers even under on-policy distillation, more strongly than under SFT and at a level comparable to off-policy distillation. Unlike SFT, where the teacher both generates the data and supervises the student, distillation lets us separate the two. With the teacher fixed, the data change how much misalignment the student inherits, but the teacher determines to which domains and tasks it transfers most strongly. Together, these results show how data and training influence the extent and patterns of emergent and subliminal misalignment.

cs.LG · cs.AI · cs.CL

版本更新 · v6arXiv:2506.14067

Online Conformal Abstention Under Adversarial Bandit Feedback

Minjae Lee, Yoonjae Jung, Sangdon Park

原文摘要

As interactive generative systems are increasingly deployed in real-world applications, their tendency to generate unreliable or false responses raises serious concerns. Conformal abstention mitigates this risk by ensuring that the system answers only when confident. However, real-world deployments typically provide only partial user feedback (e.g., thumbs up/down) on the selected response and often operate in non-stationary or adversarial environments, for which effective learning methods are largely missing. To bridge this gap, we propose ExAUL, a novel online learning framework for conformal abstention with adversarial and partial feedback. Technically, we introduce (i) a novel conversion lemma that translates the regret of any bandit algorithm into a selective risk (SR) bound, and (ii) feedback unlocking, a strategy that exploits the structure of conformal abstention to extract additional learning signals from partial feedback. We prove that ExAUL achieves a regret bound of $O(\sqrt{T \ln |H|})$, which translates into an $O(\sqrt{T})$ bound on SR violation control, matching the controllability of full-information settings despite receiving only partial feedback. While applicable to general generative tasks, we demonstrate the efficacy of ExAUL for ensuring the reliability of Large Language Models (LLMs) through empirical validation on question-answering tasks across diverse non-stationary and adversarial settings. Our results demonstrate that ExAUL robustly controls the SR while maintaining competitive answering coverage. Code and datasets are provided at https://github.com/trustml-lab/oca.

cs.LG

新论文 · v1arXiv:2611.13610

From Egocentric Human Video to Robot Manipulation in the Era of Embodied Intelligence

Yi Wang, Jingqiu Wu, Lei Yao, Zhaoshuo Li, Haoyu Zhu, Yirui Zhou, Huaiyuan Xu, Ying Zheng 等 10 位作者

原文摘要

Egocentric human video (EHV) records close-range hand-object interaction (HOI) and the temporal structure of everyday activities, offering a potentially scalable source of experience for robot manipulation. Using this experience for robot control requires recovering useful supervision without recorded robot actions and accounting for differences in morphology, viewpoint, and action space. We define Ego2Robot learning as the process of learning or adapting robot manipulation policies from EHV, organized into four learning paradigms: explicit embodiment transfer, explicit action and task supervision, egocentric video representation pretraining, and latent action and dynamics modeling. The four paradigms differ in what they derive from human video: robot-format demonstrations, interpretable supervision, visual representations, or predictive models of actions and dynamics. For each paradigm, we trace what EHV information is extracted, how robot policies use it, and when the additional data is required. Besides, we review the benchmark datasets that cover EHV, robot demonstrations, and synthetic and simulated resources. This paper links EHV data's contribution to the task information preserved and the robot-specific grounding used during training or deployment. Finally, we identify challenges in physical feasibility, agent coordination, in-context adaptation, controlled evaluation, and predictive control.

cs.RO

新论文 · v1arXiv:2611.13609

Post-OCR Correction of Mixed-Script Uzbek Text: Byte-Level ByT5 with an Empirical Homoglyph Noise Matrix

Mannon Ochilov, Oybek Narzullayev, Bahodir Zaripov

原文摘要

In Uzbek official and technical documents, a large share of OCR errors is caused not by poor image quality but by characters that are visually identical across the two alphabets in use. Because the script transition is still under way, Cyrillic and Latin appear on the same page and often within a single sentence, and in our measurements the density of homoglyph errors on such pages is close to five times higher than in single-script text. We study this problem in a setting where the corpus of engineering-geological and ICT infrastructure reports may not leave the organisation, so every stage was executed on local hardware. The proposed system combines a byte-level ByT5 model with a noise generator whose sampling weights are estimated from 359 manually classified error events, together with a Levenshtein-bounded conservative decoding rule. On a Gold set of 5,204 line pairs it lowers the character error rate from 0.0578 to 0.0369, restores 91.3% of the homoglyphs and introduces only 84 new errors.

cs.CL · cs.CV · cs.LG

版本更新 · v2arXiv:2611.09605

DKWBC: Decoupling Motion Tracking and Balancing for Humanoid Whole-Body Control

Dingkun Wei, Zehong Shen, Huaijin Pi, Xueyang Wang, Xing Zhu, Hujun Bao, Yujun Shen, Xiaowei Zhou

原文摘要

Humanoid whole-body control requires a robot to track diverse reference motions while maintaining balance under gravity, torque limits, contact switching, and external disturbances. Existing reinforcement-learning-based trackers typically learn control outputs that jointly address motion tracking and balance. However, similar reference poses can correspond to substantially different dynamic demands, requiring a unified policy to learn motion-dependent control requirements alongside balance corrections. This can complicate learning and lead to high-frequency fluctuations in control outputs. We propose DKWBC, a dynamics-informed whole-body control framework that combines explicit model-based tracking with a learned balancing residual. Inspired by the model-based and environmental-wrench contributions to joint torques in floating-base dynamics, DKWBC uses inverse-dynamics feedforward and joint PD feedback to supply tracking torques, while a residual policy learns additional joint-torque corrections for balanced execution. By explicitly accounting for motion-dependent torque demands, this structure simplifies the policy's learning task. Experiments in simulation and on real hardware demonstrate that a single learned balancing policy generalizes across motion datasets while maintaining accurate motion tracking and balance across ground support, attached payloads, and physical interaction. Project page: https://robbyant-research.github.io/DKWBC/.

cs.RO

新论文 · v1arXiv:2611.14581

When Evaluators of AI Agents Disagree: An Empirical Study of Native Scoring, LLM Judges, and Commit-Boundary Checks on tau2-bench

Sungbin Im

原文摘要

Agentic workflows, in which a large language model (LLM) calls tools to read and write data, are evaluated in two ways. Benchmarks compare the final state against a gold answer. Deployed services have none; an LLM judge reads the transcript and decides. Practice treats the two as interchangeable. This paper measures whether they are. We ran 2,160 episodes in tau2-bench's airline and retail domains and recorded every database write. We scored each episode four ways: native scoring against gold state, an LLM judge, that judge given the policy document, and the boundary evaluator, a minimal rule-based check of whether a write was expected against whether one occurred in the trace. Two annotators independently labeled 120 episodes for task correctness, policy compliance, and commitment correctness. The judge repeated its verdict 98.2% of the time but agreed with native scoring on 56.2%. In 11 episodes the agent executed a forbidden write. The judge passed 10 to 11, with or without the policy or transcript. A stronger judge still passed 6. The boundary evaluator, with no model call, caught all 11; no judge significantly exceeded it on human labels. Forbidden writes occurred in 11.1% of episodes without gates. Verification, correction, and safe-termination gates did not reduce them significantly. The traces show a candidate cause: the gates' evidence digest dropped the order status and customer authentication, so valid requests were rejected. The choice of evaluator decides the conclusion; the minimal check exposes failures the judges miss. Traces, code, prompts, and labels are released.

cs.SE · cs.AI · cs.CR

版本更新 · v4arXiv:2603.05947

LucidNFT: LR-Anchored Multi-Reward Preference Optimization for Flow-Based Real-World Super-Resolution

Song Fei, Tian Ye, Sixiang Chen, Zhaohu Xing, Jianyu Lai, Lei Zhu

原文摘要

Generative real-world image super-resolution (Real-ISR) can synthesize visually convincing details from severely degraded low-resolution (LR) inputs, but stochastic sampling can introduce semantic or structural hallucinations unsupported by the LR evidence. Sampling multiple outputs per LR input exposes trade-offs between perceptual quality and faithfulness that reinforcement learning (RL) can optimize beyond paired supervision. Effective alignment is limited by the lack of a degradation-robust LR-referenced faithfulness signal, the loss of per-objective reward contrasts when rewards are combined before normalization in Diffusion Negative-aware FineTuning (DiffusionNFT), and limited coverage of real degradations. We propose LucidNFT, a multi-reward RL framework for flow-matching Real-ISR. LucidNFT introduces LucidConsistency, a degradation-invariant LR-referenced evaluator trained with content-consistent degradation pools and hard negatives created by inpainting to detect localized hallucinations. Its decoupled reward normalization strategy normalizes each reward within each LR-conditioned rollout group before fusion, preserving per-objective contrasts across candidates. LucidLR provides a large-scale collection of real-world degraded images for RL fine-tuning. Experiments show that LucidNFT improves perceptual quality on strong flow-based Real-ISR baselines while generally maintaining LR-referenced consistency across diverse real-world scenarios. In a reward-independent human study, LucidNFT outputs were judged more faithful to the LR inputs in 72.5 percent of comparisons.

cs.CV

版本更新 · v2arXiv:2608.10380

Multiway Partition is FPT for Connectivity Functions

Tony Huynh, Eun Jung Kim, Sang-il Oum, Roohani Sharma, Marek Sokołowski

原文摘要

Let $E$ be a set of $n$ elements, $f\colon2^E\to\mathbb Z_{\ge0}$ be submodular, $t_1,\ldots,t_r\in E$ be distinct, and $k\in\mathbb N$. The Submodular Multiway Partition problem asks whether $E$ has a partition $(P_1,\ldots,P_r)$ with $t_i\in P_i$ for all $i$ and $\sum_i f(P_i)\le k$. We prove that this problem can be solved in $r^{O(k)}n^{O(1)}$ time whenever $f$ is posimodular. This covers monotone submodular functions, matroid rank functions, and nonnegative combinations of monotone and symmetric submodular functions. When $f$ is a connectivity function, we remove the dependence on the number of terminals and obtain a $k^{O(k)}n^{O(1)}$-time algorithm. The proof is entirely elementary, and is arguably the simplest proof that Edge Multiway Cut is fixed-parameter tractable. As a new consequence, our main result immediately implies that Hypergraph Multiway Partition is fixed-parameter tractable.

cs.DM · math.CO

版本更新 · v2arXiv:2611.11824

When to Trust Adaptation: Contextual Fusion for Transferable Vision-Language Prompting

Zhenyuan Chen, Wenshuo Tao, Zhuo Huang, Shuo Chen, Ming-Ming Cheng, Xiang Li

原文摘要

When a vision-language model encounters a new class or domain, how much should it trust its learned prompts? Because source adaptation can correct pretrained errors on some target images while disturbing valid predictions on others, we introduce Contextual Adaptive Fusion (CAF), which retains four prediction branches by pairing frozen and prompted visual and textual representations. To arbitrate when branches disagree on an image, CAF combines each branch's predictive confidence with how frequently the branches favor its proposed class across the unlabeled target collection, placing greater trust in predictions that are locally confident yet less dominated by collection-wide frequency bias. The fusion coefficients are learned once on source data and remain fixed at inference. Trained on ImageNet, CAF achieves 69.96% average accuracy across ten cross-dataset benchmarks (1.92 percentage points above ManiPT) and remains highly competitive across four ImageNet variants, setting a new best of 52.21% on ImageNet-S. In base-to-novel transfer across eleven datasets, CAF further reaches 79.77% novel-class accuracy and an 82.81% harmonic mean. Code is available at https://github.com/zhenyuanchenai/CAF.

cs.CV

版本更新 · v2arXiv:2611.01491

Spectral-Convolutional Factorized Transformer for Joint Multipath Parameter Estimation: Applications to mmWave Positioning and Radio SLAM

Yaya Etiabi, Wessam Ajib, Halima Elbiaze

原文摘要

Future 6G wireless networks are expected to evolve toward AI-native systems that support intelligent, environment-aware services, including positioning and mapping. Current research has already integrated AI into several aspects of communication, particularly wireless channel estimation. However, most learning-based channel estimators focus on communication-oriented objectives rather than explicitly estimating multipath parameters such as delay, angle of arrival (AoA), and angle of departure (AoD). These parameters provide essential geometric information for downstream tasks including positioning and radio simultaneous localization and mapping (SLAM). In this work, we propose a Spectral-Convolutional Factorized Transformer (SCFT) for joint multipath component (MPC) estimation from noisy complex channel estimates and physical beam-pair descriptors. SCFT combines local spectral convolution, permutation-equivariant beam-set attention, and a query-based Transformer decoder to detect and classify line-of-sight and non-line-of-sight paths and estimate their observed delays, AoAs, and AoDs. Experiments on realistic ray-tracing data show that SCFT reduces the generalized optimal sub-pattern assignment error by 13.1% compared with ResNet, the strongest competing baseline. We further demonstrate this advantage in independent snapshot positioning and radio SLAM, where SCFT reduces position RMSE by 18.5% and 65.0%, respectively, relative to ResNet. These findings highlight the importance of explicitly recovering multipath geometry, as improvements in MPC estimation can translate into substantial gains in downstream tasks.

eess.SP · cs.AI

版本更新 · v2arXiv:2609.12731

GRACE: Adaptive Concept Erasure with Geometry-Guided Retention in Diffusion Models

Qinghui Gong, Yihuai Liang, Jinshan Lai, Deepak Kumar Jain, Vitomir Štruc, Zhengchun Zhou

原文摘要

Text-to-image (T2I) diffusion models inevitably internalize sensitive or non-compliant concepts from large-scale pretraining data, necessitating post-hoc concept erasure. However, existing erasure methods often lack explicit constraints on parameter updates, leading to over-intervention and unintended semantic drift. In addition, many methods rely on manually crafted counterfactual supervision, such as surrogate prompts, which incurs substantial data construction costs that limit scalability to new concepts. To address these limitations, we propose GRACE, a structured concept erasure framework designed to enable localized and selective intervention. Specifically, we introduce a semantically weighted sensitive subspace estimation to precisely lock intervention directions, and employ lightweight subspace-constrained adapters to prevent global semantic disturbance. To eliminate the dependency on manual prompt engineering, we design an automatically decoupled safe-anchor mechanism. To mitigate semantic drift induced by excessive intervention, we introduce an energy-driven dynamic gating mechanism that adaptively controls the timing and strength of intervention at inference. Extensive experiments demonstrate that our method achieves a superior balance between erasure effectiveness and generation fidelity. Compared with the average performance of five state-of-the-art (SOTA) concept erasure methods, our method improves the fine-grained NSFW reduction rate by $17.86\%$, while reducing the macro-averaged target CLIP Score and preservation-oriented Fréchet Inception Distance (FID) by $4.55\%$ and $52.14\%$, indicating stronger concept suppression with substantially improved preservation of the original model's generative utility.

cs.CV

新论文 · v1arXiv:2611.13600

Parameterized Complexity and Geometric Approximation of Riesz Energy Subset Selection

Michael Emmerich, Soumen Atta

原文摘要

We choose $k$ of $n$ sites to minimize Riesz $s$-energy, in which a pair at distance $t$ contributes $t^{-s}$. For $s=2$, exact selection is polynomial on a line but NP-hard in the plane. Adapting the proxy lemma of Ceccarello et al. gives an efficient parameterized approximation scheme (EPAS) for fixed Euclidean dimension and fixed $s>0$. To sharpen the bound, we start with a geometric packing. Farthest-first traversal gives $2k$ centers separated by at least the last insertion distance $r$. If one site of an optimum is removed, each of the $k-1$ remaining sites excludes at most one center closer than $r/2$, leaving at least $k+1$ centers free. The new step is a discrete exchange bound: the removed site's interaction energy is at most the average it would contribute at those free centers. In the plane at $s=2$, this reduces the representative bound from $O(k^2\log(2k)/\varepsilon^2)$ for a packing estimate of total energy to $O(k\log(2k)/\varepsilon^2)$, independently of $n$. Rational planar examples show that the resulting separation scale $r/\sqrt{\log(2k)}$ is sharp. A $(1+\varepsilon)$ approximation on rational planar sites takes $O(nk+n\log n)+(C\log(2k)/\varepsilon^2)^k$ operations of polynomial bit cost. The observed number $M$ of representatives determines the search size $\binom{M}{k}$. With one million sites, $k=5$ and $\varepsilon=1/2$, the exchange grid left 252 subsets on clustered inputs and about $9.3\times10^{26}$ on uniform inputs. The method extends to metrics of bounded doubling dimension and to compact domains with effective geometric routines, including filled simple polygons. At $s=2$, exact decision on metrics with nonzero distances in $\{1,2\}$ is W[1]-complete in $k$. A gap that shrinks with $k$ rules out a deterministic EPAS on unrestricted metrics unless FPT = W[1].

cs.CG · cs.CC · cs.DS

版本更新 · v2arXiv:2611.09949

Counting Events, Not Seeing Them: Diagnosing and Repairing the Readout Failure in Video LLMs

Dongwu Qian

原文摘要

Video LLMs fail conspicuously at counting how many times an object crosses the screen, a failure attributed to missing object permanence. Using construction-grounded synthetic testbeds with validated labels, we show this attribution is incomplete: the count signal is present but unread - linear probes decode per-window presence at 93% while global count decoding collapses to a duration confound. Concurrent work reports the same representation-output dissociation for repetition counting; we show that in exit-and-reappear event counting it is repairable: two readout-directed repairs (episode-localization supervision and a discrete event-symbol side channel) raise same-object counting from 46.7% to 65.0% (oracle 66.7%) on a chance-33% benchmark across 7B-30B systems. Transfer to an independent factory costs nothing, and a measurement audit exposes five data-integrity failures that had obscured the boundary. A real-video audit (REALCOUNT, n=18, blind-audited) shows the boundary persists - best 55.6% at chance 33.3% - and symbol transfer fails zero-shot. A 1-epoch in-domain adaptation bench is negative (six levers at 28-50%), but 3 epochs reverse this for the symbol channel: with oracle timelines the reader reaches 94.4% sequence-disjoint (and 100% strict on the full set), while the deployed 2-epoch head reaches 61.1% in all three seeds (42-44), no better than a shuffled-symbol control (61.1% seed 42, 66.7% seed 43), and one clip above the unadapted reader (55.6%). The residual gap is detector evidence, not reader capacity: the same reader reaches 72.2-88.9% on GT timelines. The failure is repairable with informative timelines, but the deployed head remains the bottleneck; a frozen-feature boundary probe is weak (AUC 0.62 vs. 0.51 permutation) and non-transferring (0.77 within-clip vs. 0.62 across windows).

cs.CV · cs.CL · cs.LG

新论文 · v1arXiv:2611.13599

Unified Optimization of Traffic Signal and Lane Channelization for an Isolated Intersection Using Reinforcement Learning

Jiwon Kang, Jiyoung Kim, Hyunseok Jun, Keemin Sohn

原文摘要

Optimizing traffic signal control has been a long-lasting critical issue in traffic engineering, and a number of robust methodologies have emerged to solve the problem so far. On the other hand, the road design for lane channelization is a static problem that is difficult to solve jointly with the real-time traffic signal control. If traffic volume does not vary, it is possible that both problems can be solved in an integrated manner as a single static optimization problem. Recently, researchers have devised a mixed integer linear problem (MILP) to jointly determine the traffic signal and road width designs. However, this joint optimization problem must be solved repeatedly whenever the traffic flow pattern varies. Solving the MILP may also take impractical computing time as the problem size gets larger. We present a robust methodology for solving the joint optimization problem using reinforcement learning (RL) on a real-time basis. Abrupt changes in lane markings with an incompatible signal phase can jeopardize drivers. To sidestep such a dangerous action change, we train a model with human feedback to filter infeasible action transitions in a rule-based manner. We devised a deep Q-network that can accommodate joint actions for traffic signal and lane marking plans. To ensure the implementation of the proposed operation, a variable sign for lane markings is necessary at each stop line of an intersection. The experimental results confirm that the proposed methodology enhances the intersection performance by increasing throughput.

cs.LG

版本更新 · v3arXiv:2602.10564

SplitCom: Communication-efficient Split Federated Fine-tuning of LLMs via Temporal Compression

Tao Li, Yulin Tang, Yiyang Song, Cong Wu, Xihui Liu, Pan Li, Xianhao Chen

原文摘要

Federated fine-tuning of on-device large language models (LLMs) mitigates privacy concerns by preventing raw data sharing. However, the intensive computational and memory demands pose significant challenges for resource-constrained edge devices. To overcome these limitations, split federated learning (SFL) emerges as a promising solution that partitions the model into lightweight client-side and compute-intensive server-side sub-models, thus offloading the primary training workload to a powerful server. Nevertheless, the high-dimensional activation exchanges in SFL lead to excessive communication overhead. To overcome this, we propose SplitCom, a communication-efficient SFL framework for LLMs that exploits temporal redundancy in activations across consecutive training epochs. Inspired by video compression, the core innovation of our framework lies in selective activation uploading only when a noticeable deviation from previous epochs occurs. To balance communication efficiency and learning performance, we introduce two adaptive threshold control schemes based on 1) bang-bang control or 2) deep deterministic policy gradient (DDPG)-based reinforcement learning. Moreover, we implement dimensionality reduction techniques to alleviate client-side memory requirements. Furthermore, we extend SplitCom to the U-shape architecture, ensuring the server never accesses clients' labels. Extensive simulations and laboratory experiments demonstrate that SplitCom reduces uplink communication costs by up to 98.8\,\% in its standard configuration and total communication costs by up to 95.8\,\% in its U-shape variant without noticeably compromising model performance.

cs.NI

版本更新 · v2arXiv:2611.10441

Split and Bridge: Multimodal Generation via Diffusion Bridging

Ahmad Arrabi, Xiaohan Zhang, Xingyu Li, Safwan Wshah

原文摘要

Visual data are often expressed through multiple complementary modalities (e.g., images and segmentation masks), each capturing different aspects of the same underlying structure. Most existing multimodal methods adopt a top-down paradigm, learning a shared latent variable through joint training from scratch, which limits scalability. We propose Split and Bridge (SnB), a bottom-up framework for multimodal generation that couples pretrained unimodal diffusion models at sampling time. Rather than learning a unified latent space, SnB generates coherent multimodal samples by following a shared diffusion trajectory up to a split point and then branching into modality-specific trajectories guided by a bridge module. This design enables plug-and-play reuse of powerful pretrained models without joint training, making it scalable and easily adaptable. Across standard benchmarks (PolyMNIST, CelebAMask-HQ) and more complex settings (PIE-Bench++, ROCO), SnB achieves a strong balance between sample quality and coherence. We show that SnB naturally extends to downstream zero-shot tasks such as interleaved co-generation, image editing, and style coupling/decoupling. Our results demonstrate that bottom-up generation provides a practical framework that complements existing top-down multimodal approaches. Code is available at https://github.com/AhmadArrabi/SnB

cs.CV · cs.AI · cs.LG

版本更新 · v2arXiv:2610.07044

An exact compress-with-another threshold in a binary automaton family

Qichao Wang, Daoyu Dong

原文摘要

For a state $q$ in a synchronizing automaton, let $μ(q)$ be the minimum length of a word merging $q$ with another initial state. Szykuła asked whether $μ(q)=O(n)$ for strongly connected synchronizing automata and described a binary family attributed to Dżyga, with $n=3p$ states and a conjectured threshold $4n/3$. We specify the boundary transitions omitted from the published piecewise definition and prove that the resulting automata $\mathcal A_p$ satisfy $μ(q_0)=4p=4n/3$ for all $p\geq3$. A shortest word merging $q_0$ with another state is $(ba)^p(ab)^p$. We also construct a reset word of length $3p^2+4p-1$. Consequently, any uniform linear upper bound for $μ(q)$ on binary strongly connected synchronizing automata with corank-one letters has leading constant at least $4/3$.

cs.FL · math.CO

版本更新 · v2arXiv:2604.16682

KAIROS: Stateful, Context-Aware Power-Efficient Agentic Inference Serving

Yichao Yuan, Mosharaf Chowdhury, Nishil Talati

原文摘要

Power has become a central bottleneck for AI inference. This problem is becoming more urgent as agentic AI emerges as a major workload class, yet prior power-management techniques focus almost entirely on single-turn LLM serving. Our analysis shows that agentic serving behaves fundamentally differently: each request carries long-lived context that evolves across tool-interleaved turns, and lowering GPU frequency can push the system into a thrashing regime where memory pressure sharply worsens both performance and power efficiency. These observations show that power optimization for agentic serving requires rethinking. We present KAIROS, a context-aware power optimization system for agentic AI serving. KAIROS uses agent context as a first-class control signal to jointly manage GPU frequency, per-instance concurrency, and multi-instance request placement. This enables KAIROS to save power when memory headroom exists while avoiding thrashing and preserving performance targets. At a high level, KAIROS tracks requests at agent granularity, adapts local control to context growth and agent progress, and routes agents across instances to jointly improve power efficiency and memory stability. Evaluated across diverse software and data engineering agentic tasks, KAIROS achieves an average of 27% (up to 39.8%) power reduction while meeting the performance targets.

cs.DC · cs.AI

新论文 · v1arXiv:2611.13596

Towards Reliable VLM Judges: State-Conditional Invariance and Presentation-Aware Diagnostics

Qinan Zhang, Xinyu Wang, Qihang Jin, Tianyuan Wang, Hongbing Luo, Pengfei Li, Wenjun Lv, Yu Kang 等 9 位作者

原文摘要

Vision-language models (VLMs) are increasingly used as automated judges for multimodal and embodied tasks, yet aggregate accuracy or consistency alone does not reveal whether a judge changes its verdict for the right reason. We propose State-Conditional Invariance (SCI), a reliability principle requiring a judge to remain stable under semantics-preserving changes in presentation while responding to causal interventions that change the task outcome. We instantiate SCI in GridWM-Judge, a simulator-grounded diagnostic benchmark built from deterministic MiniGrid trajectories with aligned Full, evidence-removal, and counterfactual variants, plus presentation and representation probes. Across five diagnostic tasks, GridWM-Judge tests local perception, action-conditioned transitions, presentation stability, representation exchangeability, and Full-CF causal discrimination. Our diagnostic evaluations reveal heterogeneous failure modes, including sensitivity to evidence presentation and encoding, low activation on successful trajectories, and weak paired counterfactual discrimination. Perception-controlled analyses further indicate that visual grounding and objective recovery are substantial, model-dependent bottlenecks. These results motivate jointly reporting correctness, presentation stability, and causal sensitivity instead of relying on a single score.

cs.AI · cs.CL · cs.CV · cs.LG