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

+1865新增论文
+1473其中 AI 相关
+176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 17:12各信源进度 →

统计口径与数据新鲜度

统计 2026-11-16 17:49 — 2026-11-17 17:49(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 17:12库内论文最新更新时间:2026-11-17 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。

25,692 已收录论文21,414 AI 相关7,613 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 66 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:2611.21072

Sage S2.1 Flash Technical Report

Kun Cai, Haodong Chen, Guangran Cheng, Qingxiu Cheng, Xu Dai, Guanting Dong, Qiming Ge, Lixin Gu 等 206 位作者

原文摘要

We introduce Sage S1.5 Flash, a 320-billion-parameter general-purpose mixture-of-experts model designed to reason over scientific evidence and act in interactive environments. Sage Sci-Perception uses dedicated encoders for molecular graphs and biological sequences and represents spectra as ordered patches of measurements. Modality-specific projectors map these representations into the input space of a shared backbone, which processes them alongside text and images. To develop the model's agent capabilities, we construct verifiable environments for tool use, search, and coding, as well as specialized workflows involving machine learning experiments and scientific software. Task completion is checked against execution results, changes in environment state, or submitted artifacts. We evaluate Sage S1.5 Flash on 33 benchmarks across seven domains. In the primary comparison against five baselines, it ranks first on OfficeQA Pro (72.2), MLE-bench Lite (90.3 HumanRank), Harvey LAB, and SpreadsheetBench~2, and second on OSWorld-Verified (85.2), WebArena-Verified (67.6), and DeepResearch Bench~II (54.2). These results show particular strengths in using documentary evidence, producing verifiable workbooks, and carrying out machine learning tasks, alongside competitive performance in desktop and browser workflows that require the model to act on visual feedback.

cs.AI

版本更新 · v3arXiv:2607.11505

Proxy OPD: On-Policy Distillation with Transferable Relative Proxy Update for Large Language Models

Daocheng Fu, Rong Wu, Yu Yang, Jianbiao Mei, Licheng Wen, Xuemeng Yang, Yong Liu, Botian Shi 等 10 位作者

原文摘要

Reward-based reinforcement learning (RL) improves large language models but incurs substantial costs when repeated across targets. On-policy distillation (OPD) reuses experts by matching their distributions on target-generated trajectories, yet differing base policies can confound reward-induced updates with pre-existing policy differences, potentially distorting transferred update directions. We propose Proxy OPD (P-OPD), a post-training method that transfers reward-induced relative policy updates using the target's frozen base policy as a reference. A proxy undergoes reward-guided exploration, yielding token-level log-probability changes between its initial and optimized policies. These signals guide learning on target-generated trajectories, with calibration controlling signal absorption, enabling reuse of reward-induced improvements across targets. Experiments on Qwen3-family targets in mathematical reasoning, code generation, and instruction following show that P-OPD improves targets and outperforms conventional OPD in most settings. For 1.7B-to-8B transfer, average Math Mean@16 and Code Mean@8 scores exceed OPD by 8.3 and 3.4 points, respectively. Mechanism analyses show more faithful reward-induced supervision and controllable adaptation. Cross-scale and cross-generation experiments confirm signal transferability and reusability, while multi-domain distillation reveals synergistic gains and cross-domain interference. In the measured 4B-to-8B setting, target adaptation after proxy exploration takes approximately 15\% of direct target RL time; repeated signal reuse amortizes the one-time exploration cost across compatible models. These findings support relative policy updates as effective, transferable, and composable supervision for an ``explore once, transfer multiple times'' post-training paradigm.

cs.LG · cs.AI

版本更新 · v3arXiv:2510.23587

A Survey of Data Agents: Emerging Paradigm or Overstated Hype?

Yizhang Zhu, Liangwei Wang, Chenyu Yang, Xiaotian Lin, Boyan Li, Wei Zhou, Xinyu Liu, Zhangyang Peng 等 25 位作者

原文摘要

The rapid advancement of large language models (LLMs) has spurred the emergence of data agents, autonomous systems designed to orchestrate Data + AI ecosystems for tackling complex data-related tasks. However, the term "data agent" currently suffers from terminological ambiguity and inconsistent adoption, conflating simple query responders with sophisticated autonomous architectures. This terminological ambiguity fosters mismatched user expectations, accountability challenges, and barriers to industry growth. Inspired by the SAE J3016 standard for driving automation, this survey introduces the first systematic hierarchical taxonomy for data agents, comprising six levels that delineate and trace progressive shifts in autonomy, from manual operations (L0) to a vision of generative, fully autonomous data agents (L5), thereby clarifying capability boundaries and responsibility allocation. Through this lens, we offer a structured review of existing research arranged by increasing autonomy, encompassing specialized data agents for data management, preparation, and analysis, alongside emerging efforts toward versatile, comprehensive systems with enhanced autonomy. We further analyze critical evolutionary leaps and technical gaps for advancing data agents, especially the ongoing L2-to-L3 transition, where data agents evolve from procedural execution to autonomous orchestration. Finally, we conclude with a forward-looking roadmap, envisioning proactive, generative data agents. A curated companion repository is available at https://github.com/HKUSTDial/awesome-data-agents.

cs.DB · cs.AI

新论文 · v1arXiv:2611.19765

M-CROSS: A Multisensor Simulation Dataset for SLAM across Water--Land Transitions

Yan Dong, Tianyi Wu, Junyu Yang, Shaoqiang Qiu, Zefeng Yan, Tao Zhang, Bin Han

原文摘要

Water--land cross-domain robots require continuous localization across land, transition and underwater phases, where sensor characteristics and availability change abruptly. Existing datasets and conventional metrics are not designed to capture transition-induced degradation and failures. We present \textbf{M-CROSS}, to the best of our knowledge, the first multisensor simulation dataset and benchmark for SLAM across continuous water--land transitions. It provides synchronized LiDAR, sonar, stereo camera, IMU, DVL, and depth measurements; 6DoF pose ground truth; sensor-wise medium-state annotations; and medium-dependent sensor degradation models. We establish a transition-aware protocol comprising the Trajectory Completion Rate (TCR), Piecewise Absolute Pose Error (PW-APE), Transition End Error (TEE), and Transition Survival Period (TSP), which measure trajectory completion, single-domain accuracy, transition-induced error, and localization survival under prolonged partial immersion, respectively. Experiments with 6 representative SLAM systems show that existing methods fail to maintain both localization accuracy and trajectory continuity across the full set of water--land transitions, exhibiting transition-induced drift, trajectory discontinuities, and tracking failures. M-CROSS is publicly available https://anonymous.4open.science/r/MCross-Dataset

cs.RO

新论文 · v1arXiv:2611.19593

Learning Spatially Aware Audio-Visual Features from Videos with Mono Audio

Mengyu Yang, You Zhang, Vishnu Raj, Olha Townsend, James Hays, Andrea Fanelli, Gautam Bhattacharya

原文摘要

Comprehensive audio-visual perception requires both semantic and spatial reasoning. Existing spatial models suffer from poor generalization due to data scarcity, while semantic-only models fail to capture spatial characteristics completely. We introduce a model capable of jointly reasoning over binaural audio and vision on semantic and spatial tasks with improved generalization. To overcome the shortage in real-world data, we develop an automatic pipeline that localizes the sound source in a video, estimates its relative 3D trajectory, and renders binaural audio. We then use this to spatialize VGGSound, allowing us to train at a much larger scale. We demonstrate improved performance and generalization on spatial reasoning tasks while still remaining competitive on semantic tasks compared to semantic-only models.

cs.CV

版本更新 · v3arXiv:2611.13223

Deletion Sensitivity Is a Property of the Model: A Three-Condition Measurement over 72 Valid Experimental Cells

Yiran Li, Hanzhang Jia, Liheng Zeng, Zhiyu Yang, Yi Gao

原文摘要

Reasoning models write a trace at every turn of a multi-turn tool session, and later turns read it. Upstream providers hold different retention constraints, so a request forwarded across issuers arrives with its history deleted, and its task cost remains unmeasured. This paper reports a direct, blind, multi-seed measurement of full history deletion in multi-turn tool calling, through three rendering conditions, a calibrated set of 72 valid experimental cells, a multi-seed protocol, and the official checker. The same deletion separates the models by direction and by magnitude, one class losing tens of points, another gaining, and a third unchanged, with the endpoint ordering reproduced across seeds and task subsets modulating the magnitude and, near zero, the sign. A re-dosed model preserves its success across the full deletion range, so the reading belongs to the model rather than the dose. Deletion sensitivity is a model-level property under an operational definition, measured under a fixed task family, scorer, and protocol, and a retrained model is a new model that a two-cell screen, one retention run and one deletion run, re-reads for the cost of two runs. Deletion sensitivity is thus a measurable property of the model, and the reading must enter routing as prior information rather than persist as an invisible default cost.

cs.CL · cs.AI · cs.SE

新论文 · v1arXiv:2611.16419

I See What You Want: Decoding Hierarchical Operator Intent from Visual Teleoperation History

Xiangyi Jia, Kaixian Qu, Shengzhi Wang, Chenyu Yang, Robert Katzschmann, Marco Hutter

原文摘要

Vision-language models (VLMs) have become a powerful foundation for robotic scene understanding, task planning, and language-conditioned control. A natural question follows: can these models also decode what a human operator intends while teleoperating a robot, without any explicit instruction? We formulate this as hierarchical intent prediction, jointly inferring the ongoing atomic action and the broader mission it serves from a short visual teleoperation history. To study this, we introduce VI AMO (Vision Intention Action and Mission Ongoing), a dataset of 6,600 episodes from eleven robot sources with 13,653 temporally aligned action annotations and 9,702 mission annotations. Large proprietary models struggle with this problem: on held-out episodes, GPT-4o reaches Sentence-BERT similarities of only 0.563 and 0.470 for action and mission, and GPT-5.6 Luna achieves only 0.581 and 0.554. Fine-tuning a VLM on VI AMO yields our Vision-Intention Model (VIM), which raises these scores to 0.825 and 0.821 and still reaches 0.726 and 0.717 on the 35.4% of test actions with both labels absent from training. Finally, as a downstream application, VIM's predicted mission enables a vision-language-action (VLA) policy to take over from the operator mid-task in a real-world environment unseen during VIM fine-tuning, reducing human teleoperation time by 34.6% relative to GPT-4o and 39.8% relative to GPT-5.6 Luna, while cutting intent-prediction latency by roughly 70% on average.

cs.RO · cs.CV

新论文 · v1arXiv:2611.14848

Tideway: Phase-Aware Tiered Memory Management for MoE Inference on Consumer Devices

Yifei Wang, Tsun-yu Yang, Yechen Xu, Hancheng Ye, Qinsi Wang, Baijun Zhou, Tingjun Chen, Hai "Helen" Li 等 10 位作者

原文摘要

Large language model (LLM) agents are rapidly becoming a dominant paradigm for AI applications. Recently, there is a growing demand to deploy large-scale Mixture-of-Experts (MoE) models locally on consumer PCs to balance data privacy, API costs, and inference efficiency. However, unlike datacenter servers with abundant DRAM, consumer devices are tightly constrained in both GPU VRAM and host RAM, forcing massive MoE weights to spill across VRAM, host RAM, and SSD. Under such spilling, we observe that existing systems fail to fully utilize this memory hierarchy, treating the tiers as independent caches and thrashing in inference. We present TIDEWAY, a system for efficient inference of large MoE models on memory-constrained consumer devices. Our core insight is to manage host RAM and GPU VRAM as a unified memory pool. Within this pool, TIDEWAY couples phase-aware memory layouts, which adapt to the distinct access patterns of prefill and decode, with an execution engine that optimizes CUDA Graph execution under expert paging. Our evaluations on representative consumer hardware with frontier MoE models demonstrate that TIDEWAY significantly outperforms state-of-the-art systems. On average, TIDEWAY achieves speedups of 1.97x in end-to-end latency and 2.40x in Time-to-First-Token (TTFT), along with 1.61x higher decode throughput against the fastest baseline. Our source code is publicly available at https://github.com/Afterglow231/Tideway .

cs.OS

版本更新 · v3arXiv:2604.26963

MARS: Towards High Goodput and Low KV Thrashing in Agent Serving under GPU Memory Constraints

Yifei Wang, Tsun-yu Yang, Hancheng Ye, Yechen Xu, Cong Guo, Chiyue Wei, Qinsi Wang, Dongting Li 等 12 位作者

原文摘要

Self-hosted LLM agents offer greater control over sensitive data and model customization while avoiding per-request API fees. However, serving concurrent, multi-turn workflows within limited GPU memory makes it difficult to achieve high workflow-level goodput. In particular, as workflows accumulate context, their reusable key-value (KV) caches can easily outgrow GPU capacity, triggering repeated eviction and recovery, known as KV cache thrashing. This process consumes resources and delays workflow completion. Although existing approaches propose various methods, our motivation study shows that the evaluated baselines still incur substantial thrashing overhead or fail to address the workflow-level goodput issue. We present MARS, an agent serving system that coordinates workflow admission and in-engine scheduling to achieve high workflow-level goodput on memory-constrained devices while avoiding KV cache thrashing. Specifically, outside the engine, an admission controller selects root workflows using initial prefix footprint estimates and live KV availability, adapting admissions to memory pressure. Inside, an agent-centric scheduler prioritizes turn completion and coordinates prefill allocation with KV reclamation to reduce disruptive evictions. Across two models and three GPU configurations, MARS achieves 1.59x-7.56x mean end-to-end latency speedups over the fastest baseline. Integrated as the serving backend for OpenCode, Codex, and OpenHands, MARS further improves mean workflow completion time by 1.17x-1.87x. Our source code is publicly available at https://github.com/Afterglow231/MARS_preview .

cs.OS · cs.DC · cs.LG · cs.MA

版本更新 · v2arXiv:2611.10693

Machine Hearing: Can Audio LLMs Generalize Beyond Human Auditory Perception?

Liwei Liu, Jun Tao, Xingyuan Li, Xuenan Xu, Chen Zhang, Junyi Wang, Boya Zhang, Xiaoyu Yang 等 11 位作者

原文摘要

Multimodal LLMs have made remarkable progress in perceiving the physical world, while their perceptual scope remains largely human-centric, focusing on what humans can directly see or hear. This progress raises the prospect of perception beyond human limits, such as extending hearing to sounds that are inaudible to humans. More broadly, the physical world is full of waves, vibrations, and oscillations, only a small fraction of which are directly audible to humans. This invites the extension of multimodal LLMs toward a broader form of machine hearing, which remains largely underexplored. To investigate this direction, we introduce the Machine Hearing Benchmark (MHB), comprising 6,128 instances across 13 tasks and seven domains. MHB spans highly heterogeneous input regimes, with sampling rates from approximately $10^{-5}$ to $10^6$ Hz, sequence lengths from $10^1$ to $10^6$ samples, and both single- and multi-channel inputs. Evaluations of current open-source and proprietary Audio LLMs (ALLMs) reveal limited zero-shot transfer to machine hearing. Supervised adaptation on machine-hearing data alone does not close this gap, suggesting that the challenge extends beyond data exposure to a systematic mismatch between existing audio front ends and machine-hearing inputs. To address this mismatch, we introduce MachEar, a lightweight encoder extension that complements the pretrained Audio LLM for machine-hearing inputs. The results show that MachEar equips a pretrained Audio LLM with stronger machine-hearing capabilities while largely preserving its existing audio understanding performance.

cs.SD · cs.CL

版本更新 · v2arXiv:2609.17769

Prior-Aided Masked Vector Quantization CSI Feedback for FDD Massive MIMO Systems

Yi Song, Tianyu Yang, Kangda Zhi, Shuangyang Li, Giuseppe Caire

原文摘要

Downlink channel state information (CSI) feedback is a key bottleneck in frequency-division duplex (FDD) massive MIMO systems, as the user equipment (UE) must convey its estimated channel to the base station (BS) over a limited uplink (UL) budget. To improve CSI reconstruction accuracy under tight feedback constraints, we propose prior-aided masked vector quantization (PM-VQ), a learning-based separate source--channel coding (SSCC) feedback scheme conditioned on the average angle--delay power map---a compact representation of the channel second-order statistics available at both the UE and the BS. In PM-VQ, a prior-aided encoder maps the CSI to latent tokens, and a spatially-adaptive masking module (SAMM) scores and selects the most informative tokens within the feedback budget. The selected tokens are vector-quantized and fed back together with their positions, while an adaptive de-masking module (ADM) completes the latent representation at the BS before prior-conditioned decoding. To support variable-rate compression, a single model is trained over a range of selected-token counts, enabling operation across multiple feedback dimensions without retraining. We evaluate PM-VQ against three representative baselines on a Sionna-generated 3GPP TR~38.901 UMa dataset, focusing on the most challenging diffuse regime where channel energy is spread across many angle--delay coefficients. Simulation results show that PM-VQ achieves the lowest NMSE across all tested SNR levels and feedback dimensions in this regime. Moreover, the angle--delay power-map prior remains beneficial even when estimated from only a few channel realizations.

cs.IT

版本更新 · v2arXiv:2605.15721

Contexting as Recommendation: Evolutionary Collaborative Filtering for Context Engineering

Jiachen Zhu, Zhuoying Ou, Congmin Zheng, Yuxiang Chen, Zeyu Zheng, Rong Shan, Lingyu Yang, Lionel Z. Wang 等 12 位作者

原文摘要

Large Language Models (LLMs) are highly sensitive to their input contexts, motivating the development of automated context engineering. However, existing methods predominantly treat this as a global search problem, seeking a single context strategy that maximizes average performance across a dataset. This restrictive assumption overlooks the fact that different inputs often require distinct guidance, leaving substantial instance-level performance gains untapped. In this paper, we propose a paradigm shift by formulating context engineering as a recommendation problem. We introduce \textbf{Neural Collaborative Context Engineering (NCCE)}, a framework that transitions optimization from a static global search to dynamic, instance-wise routing. NCCE first bootstraps a diverse catalog of anchor contexts and then employs a novel \textbf{Context-CF Co-Evolution} mechanism. This stage establishes a synergistic feedback loop: a lightweight Neural Collaborative Filtering (NCF) model learns instance-context preferences to guide the generation of specialized context variants, while the newly evaluated contexts continuously refine the NCF model's understanding of latent preferences. At inference time, the trained NCF model acts as a context router, dynamically assigning the most suitable context strategy to each unseen instance. Theoretical Proofs and comprehensive experiments demonstrate that by matching individual inputs with their optimal contexts, NCCE significantly improves task accuracy, highlighting the critical importance of personalization in LLM context engineering.

cs.CL

版本更新 · v3arXiv:2611.03412

DeliverBench: Benchmarking AI Agents on Delivering Realistic Multi-Stage Professional Projects

Yumou Liu, Mingju Chen, Zhoufutu Wen, Qianyu Yang, Zijie Wang, Sihan Chen, Yeqian Liu, Chenghao Yang 等 24 位作者

原文摘要

LLM agents are judged by what they deliver, and in professional work what gets delivered is a project, not a task. A project pays off only if the whole package holds together, and one weak part can limit the value of the rest. Yet existing benchmarks mostly test agents on one task at a time, leaving a gap between the capability we measure and the outcome that makes agents useful. To address this gap, we introduce DeliverBench, which tests whether agents can carry a project from brief to final delivery. Reconstructed from real materials with 13 domain experts, its 11 projects span settings such as financial due diligence, legal contract drafting, and drug safety reporting. They comprise 150 subtasks, with 2-15 subsequent releases of new materials and client feedback per project. Later subtasks build on the agent's own earlier deliverables, so unresolved errors can carry into downstream work. Expert rubrics assess each subtask submission and the final delivery package. Across seven model-harness systems, the best scores 65.0/100 on final delivery and passes 50.0% of subtasks at a 90-point threshold. Together, these results suggest that task-level capability does not simply compose into delivery; delivery also depends on keeping the work coherent as it evolves.

cs.AI

新论文 · v1arXiv:2611.12382

AutoVLA-D: Learn a Trajectory-Geometric Condition from Reasoning for Autonomous Driving VLA

Yuguang Yang, Zhewen Tan, Canyu Chen, Linlin Yang, Cheng Chi, Chunyang Liu, Kehua Sheng, Jinyu Yang 等 12 位作者

原文摘要

Autoregressive VLA planning usually predicts continuous trajectories as the final output, but driving instructions often describe actions rather than ego states, which weakens direct conditioning on the final continuous trajectory. We address this gap with TrajCond, which introduces a discrete trajectory-geometric condition for VLA planning. Specifically, we represent TrajCond as a 1D-Brohan differential action trajectory and learn it with a two-stage training framework before final continuous trajectory prediction. With the simultaneous autoregressive generation of discrete and numeric trajectories, we name our method AutoVLA-D (Double). Experiments on NAVSIM navtest, nuScenes, and NeuroNCAP show that discrete trajectories are weaker as direct outputs but more effective as TrajCond, consistently improving final continuous trajectory prediction across closed-loop and open-loop driving benchmarks. Code is available at https://github.com/ShivaTeam/AutoVLA-D.

cs.RO · cs.CV

版本更新 · v3arXiv:2606.15912

On-Policy Distillation with Curriculum Turn-level Guidance for Multi-turn Agents

Gengsheng Li, Mao Zheng, Mingyang Song, Ruiqi Liu, Tianyu Yang, Jie Sun, Qiyong Zhong, Haiyun Guo 等 11 位作者

原文摘要

Multi-turn agents that plan, invoke tools, and interact with environments offer a promising paradigm for solving complex tasks, yet their capabilities typically rely on very large models whose inference cost is prohibitive in practice. On-Policy Distillation (OPD) is a natural recipe for transferring such capabilities to smaller students, but we find that it suffers a characteristic failure mode in this setting: small student errors compound across turns and push the trajectory out of the teacher's familiar state distribution, so the teacher's supervision becomes least reliable precisely where the student needs it most. We propose Guided On-Policy Distillation (Guided-OPD), a simple yet effective algorithm that mixes teacher- and student-generated turns within each rollout and schedules the teacher's intervention probability along a curriculum that decays to zero. Strong guidance keeps early trajectories close to the teacher distribution and is then gradually withdrawn to recover the purely on-policy regime used at inference. On ALFWorld, ScienceWorld, and WebShop, distilling Qwen3 students from a Qwen3-30B-A3B teacher, Guided-OPD yields average relative gains of 21.1\% in Score and 25.5\% in Success Rate over vanilla OPD, with larger gains on smaller students.

cs.LG · cs.AI

版本更新 · v3arXiv:2512.14329

Wearable-informed generative digital avatars predict task-conditioned post-stroke locomotion

Yanning Dai, Chenyu Tang, Ruizhi Zhang, Wenyu Yang, Yilan Zhang, Yuhui Wang, Junliang Chen, Xuhang Chen 等 19 位作者

原文摘要

Dynamic prediction of locomotor capacity after stroke could enable more individualized rehabilitation, yet current assessments largely provide static impairment scores and do not indicate whether patients can perform specific tasks such as slope walking or stair climbing. Here, we present a wearable-informed data-physics hybrid generative framework that reconstructs a stroke survivor's locomotor control from wearable inertial sensing and predicts task-conditioned post-stroke locomotion in new environments. From a single 20 m level-ground walking trial recorded by five IMUs, the framework personalizes a physics-based digital avatar using a healthy-motion prior and hybrid imitation learning, generating dynamically feasible, patient-specific movements for inclined walking and stair negotiation. Across 11 stroke inpatients, predicted postures achieved approximately 95% of the practically attainable accuracy for slope ascent and 80% for stair climbing, respectively, demonstrating reliable task-specific gait prediction. In a multicenter randomized pilot feasibility study (n = 21; 28 days), access to scenario-specific locomotion predictions to support task selection and difficulty titration was associated with larger gains in Fugl-Meyer lower-extremity scores than the control group (mean change 6.0 vs 3.7 points; p < 0.05). These results suggest that wearable-informed generative digital avatars may augment individualized gait rehabilitation planning and provide a pathway toward dynamically personalized post-stroke motor recovery strategies.

cs.CE · cs.AI

版本更新 · v2arXiv:2605.21699

X-Token: Projection-Guided Cross-Tokenizer Knowledge Distillation

Sharath Turuvekere Sreenivas, Adithyakrishna Venkatesh Hanasoge, Mingyu Yang, Ali Taghibakhshi, Saurav Muralidharan, Ashwath Aithal, Pavlo Molchanov

原文摘要

Cross-tokenizer knowledge distillation allows a student to learn from teachers with incompatible vocabularies while preserving its tokenizer and inference architecture. Existing logit-based methods either supervise only the observed token or match full distributions through strict token partitions and heuristic ranking. We identify two limitations: an uncommon-token failure, where task-critical tokens outside the shared vocabulary are suppressed, and over-conservative matching, where exact string equality discards useful correspondences. We propose X-Token, which addresses both through a sparse projection initialized from tokenizer-level string rules and optionally refined during distillation. P-KL removes the partition and projects the student distribution into the teacher vocabulary; H-KL retains the hybrid loss while relaxing matching through the projection. A token-coverage audit guides the choice between them, and both extend naturally to heterogeneous teachers. Across Llama, OLMo, and Gemma students, X-Token achieves the highest average accuracy in all four teacher-student settings among all evaluated cross-tokenizer methods. Complementary teachers provide a further 1.30-point gain, and both variants are faster than the evaluated cross-tokenizer baselines in matched profiling.

cs.LG · cs.CL

新论文 · v1arXiv:2611.10356

SepLRM: A Foundation Model Paradigm for Recommendation

Di Chen, Xijun Xiao, Qin Ren, Weijiang Lai, Zhijian Pan, Zhentao Pan, Pengfei Nie, Cheng Yuan 等 26 位作者

原文摘要

Large recommendation models (LRMs) improve recommendation performance by scaling behavior sequence length, model capacity, and training compute. However, the return on investment (ROI) of continued scaling deteriorates as marginal gains diminish and long-sequence encoding costs rise, especially when this computation is repeatedly executed by distinct models for diverse recommendation stages and business scenarios. To address this issue, we propose SepLRM, a recommendation foundation model (RFM) paradigm that identifies reusable sequence encoding and reorganizes multiple end-to-end (E2E) LRMs into one RFM and multiple independently optimized vertical models (VMs). The RFM adopts an encoder-decoder architecture and exports its final-layer encoder state as the shared key-value (KV) cache that is independent of downstream task-offered candidates. Unlike candidate-specific logits or aggregated user embeddings, the KV cache preserves fine-grained behavioral information. Separation at this native KV interface also allows each VM to construct candidate-conditioned queries while retaining the same query-KV interaction as its E2E counterpart, enabling high-fidelity capability transfer. To ensure both transfer effectiveness and system efficiency, we co-design cache compression, the training and serving workflows, and LG-Store, a distributed storage system for KV caches. Experiments on three large-scale datasets show that SepLRM achieves state-of-the-art offline performance with a transfer rate of approximately 100%. It also demonstrates strong transferability across model iterations, recommendation stages, and business scenarios. Deployment in two large-scale industrial scenarios substantially improves the ROI of sequence scaling and delivers consistent gains in online A/B tests at ByteDance.

cs.IR

版本更新 · v3arXiv:2601.09920

SyncTwin: Fast Digital Twin Construction and Synchronization for Safe Robotic Manipulation

Ruopeng Huang, Boyu Yang, Wenlong Gui, Jeremy Morgan, Erdem Biyik, Jiachen Li

原文摘要

Accurate and safe robotic manipulation under dynamic and visually occluded conditions remains a core challenge in real-world deployment. We introduce SyncTwin, a novel digital twin framework that unifies fast 3D scene reconstruction and real-to-sim synchronization for robust and safety-aware robotic manipulation in such environments. In the offline stage, we employ VGGT to rapidly reconstruct object-level 3D assets from RGB images, forming a reusable geometry library. During execution, SyncTwin continuously synchronizes the digital twin by tracking real-world object states and uses stored assets to complete partially observed geometry. The synchronized twin enables motion planners to compute collision-aware robot trajectories in simulation, with execution on the real robot closing the real-to-sim-to-real loop. Experiments in dynamic and occluded scenes show that SyncTwin improves manipulation performance and motion safety, demonstrating the effectiveness of digital twin synchronization for real-world robotic execution.

cs.RO

版本更新 · v2arXiv:2611.07847

Computation, not weight access: large language models as auditors of clinical prediction networks

Zhiyu Liu, Xinyu Yang

原文摘要

Frontier large language models (LLMs) have been reported to read, edit and write the weights of small neural networks, raising the prospect of auditing clinical prediction models from their files alone. Public demonstrations, however, mixed reading raw weights with running code. We separated these conditions in internally pre-specified simulation experiments. Tanh networks of 105 to 38,401 weights were trained on simulated cohorts of patients with subarachnoid haemorrhage, each with a planted relationship or none, and one network was trained on labels that under-recorded deaths in a subgroup. GPT-6 Astra and four other GPT models answered single prompts under a default or an explicit system instruction, either with the weights in the prompt and no tools, or with a calculation-only tool that exposed the weights (white box) or only the network's predictions (black box). Without tools, Astra recovered planted relationships in 353-weight networks but became unreliable at 3,457 weights, and under the default instruction it declined 17 of 21 audit and repair tasks as unidentifiable; the explicit instruction removed these refusals. With the tool, it recovered the planted relationship in all 39 runs on networks of up to 38,401 weights, raised no false alarm in 21 runs on null networks, and found the biased subgroup in 6 of 6 runs. Black-box access performed as well as white-box access, and other GPT-6 models read as accurately. A well-designed script matched the comparator models on the audit and the best LLM repair. For auditing these networks, what mattered was computation and elicitation, not access to the weights.

cs.LG · cs.AI

版本更新 · v2arXiv:2602.04144

OMG-Agent: Toward Robust Missing Modality Generation with Decoupled Coarse-to-Fine Agentic Workflows

Ruiting Dai, Zheyu Wang, Haoyu Yang, Yihan Liu, Chengzhi Wang, Zekun Zhang, Zishan Huang, Jiaman Cen 等 9 位作者

原文摘要

Data incompleteness severely impedes the reliability of multimodal systems. Existing reconstruction methods face distinct bottlenecks: conventional parametric/generative models are prone to hallucinations due to over-reliance on internal memory, while retrieval-augmented frameworks struggle with retrieval rigidity. Critically, these end-to-end architectures are fundamentally constrained by Semantic-Detail Entanglement -- a structural conflict between logical reasoning and signal synthesis that compromises fidelity. In this paper, we present \textbf{\underline{O}}mni-\textbf{\underline{M}}odality \textbf{\underline{G}}eneration Agent (\textbf{OMG-Agent}), a novel framework that shifts the paradigm from static mapping to a dynamic coarse-to-fine Agentic Workflow. By mimicking a \textit{deliberate-then-act} cognitive process, OMG-Agent explicitly decouples the task into three synergistic stages: (1) an MLLM-driven Semantic Planner that resolves input ambiguity via Progressive Contextual Reasoning, creating a deterministic structured semantic plan; (2) a non-parametric Evidence Retriever that grounds abstract semantics in external knowledge; and (3) a Retrieval-Injected Executor that utilizes retrieved evidence as flexible feature prompts to overcome rigidity and synthesize high-fidelity details. Extensive experiments on multiple benchmarks demonstrate that OMG-Agent consistently surpasses state-of-the-art methods, maintaining robustness under extreme missingness, e.g., a $2.6$-point gain on CMU-MOSI at $70$\% missing rates.

cs.AI · cs.LG

版本更新 · v2arXiv:2611.08080

Spatiotemporal Dynamics-Aware Industrial Anomaly Detection and Reasoning

S-Wei Chen, Siangyu Yang, Yi-Chen Jan, Chun-Yi Lee, Shao-Yuan Lo

原文摘要

Industrial Anomaly Detection (IAD) is a long-established problem that aims to detect defects in industrial products during manufacturing for quality control. Most existing studies focus on static visual patterns as the basis for detection. However, in real manufacturing environments, many anomalies are dynamics-based, such as irregular movements that cannot be perceived from static cues alone. We observe that even recent Multimodal Large Language Model (MLLM)-based approaches fall short in identifying such dynamic anomalies due to limited temporal modeling capability. To address this, we introduce MUSTARD: MUltimodal SpatioTemporal Anomaly Reasoning and Detection, a framework that equips MLLMs with fine-grained spatiotemporal dynamics awareness for detecting and reasoning about dynamics-based defects. MUSTARD introduces optical-flow-guided keyframe selection and augmented modalities to capture critical motion irregularities through local motion patterns and global dynamic evolution. Its normal pattern-based reasoning pipeline requires only a few normal reference samples as the detection basis, avoiding any model updates. This enables efficient adaptation across various products and provides high decision interpretability. MUSTARD is the first MLLM-based framework for dynamics-based IAD. Experiments demonstrate its superior performance in detecting spatiotemporal anomalies and providing plausible explanations. Our code is available at https://github.com/racercode/Mustard

cs.CV

版本更新 · v2arXiv:2606.30015

参数化技能

Parametric Skills

Xuan Zhao, Haonan He, Qingyu Yang, Minglei Li, Zelin Tan, Bo Wan, Jingqi Ye, Chen Zhang 等 10 位作者

AI 速读 基于摘要

提出 ParametricSkills 框架,在测试时将文本技能转化为 LoRA 适配器,实现无需占用上下文的技能全生命周期管理。

cs.CL

查看原文摘要

For LLMs, skills are textual recipes encoding mature problem-solving experience. Despite widespread deployment, their utility is limited by the model's ability to comprehend and follow skill instructions. To address this limitation, we propose ParametricSkills, a framework that parameterizes textual skills at test time, enabling context-free management of the full skill lifecycle. Specifically, we first construct a large-scale and high-quality skill collection, and then synthesize single-turn and multi-turn skill-exploitation trajectories. Using this data, we then train a hypernetwork that parameterizes both the content and the exploitation methodologies of skills by converting textual skills into LoRA adapters. To facilitate the real-world application of our framework, we further align the hypernetwork on complex, long-horizon trajectories collected with OpenCode, enabling the models to operate reliably in realistic interactive workflows with ParametricSkills. Across evaluations, ParametricSkills effectively parameterizes skills, yielding substantial gains in task completion. More notably, as parametric skills accumulate as adapters rather than crowding the context, they open a preliminary but promising path toward continual learning at test time.

新论文 · v1arXiv:2611.08367

CascadeFilter:联合阈值选择的高成本效益语义过滤

CascadeFilter: Cost-Efficient Semantic Filtering with Joint Threshold Selection

Yanji He, Tao Zhong, Hongyang Chen, Xiaobo Zhang, Mingyu Yang, Wei Wang

AI 速读 基于摘要

提出三阶段语义过滤方法 CascadeFilter,将阈值选择建模为带 oracle 相对误差约束的 token 成本最小化问题,并在预定义网格中联合搜索小模型阈值。

cs.DB

查看原文摘要

Semantic filtering selects records that satisfy a natural-language predicate and is an important operation in data analytics based on large language models (LLMs). Executing an expensive model on every document incurs substantial inference cost. Existing approaches reduce this cost by accepting predictions from cheaper models and forwarding uncertain records to an oracle. Task Cascades further varies document portions and operations, and provides an accuracy-guaranteed extension. However, its threshold calibration follows a prescribed adjustment sequence after cascade assembly, rather than minimizing cost over alternative combinations of stage thresholds. Motivated by this, we propose CascadeFilter, a three-stage semantic filtering method that jointly searches the thresholds for a small language model operating on extracted evidence and full documents, with unresolved records forwarded to the oracle. For a fixed cascade, we formulate threshold selection as minimizing token-aware cost subject to an oracle-relative error constraint, and select the least-cost certified configuration in a pre-specified grid. A finite-population calibration procedure bounds oracle-relative error after selection, using fixed-sequence tests on threshold tracks to allocate the multiple-testing budget. We further combine lexical and dense retrieval, reranking, and neighborhood-supported confidence to improve evidence extraction and routing. These components enhance the cascade while leaving its certification interface unchanged. Across six datasets and 13 evaluation settings, the reported results have the highest benchmark accuracy and macro-averaged F1 score in 12 settings each. These comparisons describe the reported operating points; the accuracy guarantee concerns agreement with the oracle.

新论文 · v1arXiv:2611.08318

学习任务还是学习验证器?重新审视具身自我改进中的评判标准暴露

Learning the Task or Learning the Verifier? Rethinking Criteria Exposure in Embodied Self-Improvement

Wenbo Huang, Jianxin Bi, Shunyu Yang, Ruiyang Si, Rui Ni, Qiang Wang, Shulong Jiang, Yintianrun Zhang 等 13 位作者

AI 速读 基于摘要

作者通过受控实验质疑具身自我改进智能体在基准测试中暴露成功评判标准(Criteria Exposure)的做法,指出其可能诱导代码过拟合验证器而非真正完成任务。

cs.AI · cs.RO

查看原文摘要

Self-improving embodied agents are increasingly being developed and compared on robot simulation benchmarks. Several recent works expose the benchmark's success criteria to the learning agent, a practice we call Criteria Exposure (CE). These criteria, originally written to verify policies trained by imitation learning, often demand more than the corresponding language instructions, so a score from such a verifier may not faithfully reflect whether the instructed task was completed, especially for robot-control programs written by an agent. As embodied coding agents report ever higher benchmark scores, a natural question arises: when a coding agent is exposed to the benchmark's criteria, does it generate code that better performs the instructed task, or code that simply overfits to the verification criteria? We conduct controlled experiments with a self-improving agent with and without CE, and find that with CE the agent quickly produces code with a higher success rate under the given criteria, but this advantage disappears when the criteria are changed slightly without departing from the original instructions. More importantly, the skills the agent develops under CE transfer poorly across tasks. We further propose instruction-faithful criteria and show that they better reflect the performance of agent-generated programs.

新论文 · v1arXiv:2611.07665

CoGenRec: A Unified Generative Framework for Complementary Vision Reconstruction

Yapeng Meng, Yuguo Chen, Xiangru Chen, Shaojun Xu, Xuezhen Wang, Lijian Wang, Zheyu Yang, Yihan Lin 等 10 位作者

原文摘要

Capturing scenes with extreme motion and high dynamic range is fundamentally constrained by trade-offs among temporal resolution, spatial detail, signal-to-noise ratio, and dynamic range under practical bandwidth, power, and quantum-efficiency limits. Conventional frame-based cameras therefore suffer from temporal undersampling, exposure-induced motion blur, and sensor saturation. The Complementary Vision Sensor (CVS) addresses these limitations by decomposing visual information across two complementary pathways, a spatially dense, temporally sparse, long-exposure RGB pathway and a temporally dense, spatially sparse, short-exposure patiotemporal difference pathway. Their complementary measurements provide appearance, structural, and motion cues across a wide dynamic range that enable reconstruction beyond conventional imaging. We present CoGenRec, a complementarity-driven generative reconstruction framework that unifies video deblurring, high-temporal-resolution frame interpolation, and HDR fusion. CoGenRec encodes complementary measurements into multi-scale spatiotemporal features using exposure-aware appearance modeling and bidirectional temporal propagation. These features constrain a video diffusion model and are injected into the VAE decoder to recover fine structures lost during latent-space compression, combining the scene fidelity of physical measurements with the appearance recovery capabilities of learned video priors. On SportsSloMo-CVS, CoGenRec improves PSNR over the strongest general-purpose baseline by 4.40 dB for deblurring, 3.89 dB for frame interpolation, and 5.04 dB for HDR fusion, respectively. It also achieves strong perceptual quality on 129 unseen scenes captured with real CVS. These results establish a unified approach to advance high-quality reconstruction under challenging motion and illumination conditions. Project page: https://cogenrec.github.io/.

cs.CV

新论文 · v1arXiv:2611.07542

FlashOffloading: Low-Latency KV Cache Offloading for Long-Context LLM Serving

Xuan Wang, Zhuanhao Wu, Tingzhu Fang, Xushi Li, Sicheng Zhang, Zhenyu Yang, Feng Liu, Zhaohui Fan 等 20 位作者

原文摘要

In large language model (LLM) serving systems, the key-value (KV) cache grows linearly with sequence length and concurrency, so long-context inference easily hits the high-bandwidth memory (HBM) capacity wall and results in low concurrency. Offloading the KV cache to host dynamic random-access memory (DRAM) frees HBM space and raises concurrency, but the additional cache management and host-to-device (H2D) load overhead inflates inference latency, higher concurrency fails to translate into effective throughput. We present FlashOffloading, a host-DRAM-centric inference architecture that fully offloads the KV cache in both the prefill and decode phases. FlashOffloading builds rolling HBM cache for prefill, a hot cache for decode, a reverse sparse index with zero extra miss-lookup overhead, and a compute--load overlapped kernel, optimizing KV offloading latency across the entire pipeline and delivering throughput gains under a time-per-output-token (TPOT) service-level objective (SLO). In experiments deploying GLM-5.2 with multi-token prediction (MTP3) speculative decoding on an Ascend910C prefill--decode (PD)-disaggregated platform, FlashOffloading improves prefill throughput by up to 1.38\(\times\) and decode throughput by up to 1.91\(\times\) under a 30 ms TPOT constraint. FlashOffloading is integrated into an open-source serving framework: https://github.com/vllm-project/vllm-ascend/tree/releases/v0.30.0.

cs.AI

新论文 · v1arXiv:2611.07232

Recursive Sparse Attention: Million-Token Length Extrapolation at Logarithmic Cost

Ruitian Wang, Xinyu Yang, Minshen Zhang, Hao Zhang

原文摘要

At context lengths of about 1M tokens, dense attention is costly in prefill and decoding, and training typically relies on long documents, which become scarcer as the target length increases. Most sparse attention methods score all preceding blocks or tokens, so their selection cost per query is $O(L)$ at context length $L$; hierarchical selection reduces it to $O(\log L)$, but prior trainable hierarchical methods report accuracy only within their training length. We propose RecSA (Recursive Sparse Attention), which builds a tree over the KV cache in which each node summarizes its subtree with one key and one mean-pooled value. The keys come from one pooler per layer, shared by all levels and distilled from dense attention. Each level above the tokens selects at most $k$ candidates, so each query has $O(\log L)$ selection cost and reads $O(\log L)$ entries; the summaries of unselected candidates share one softmax with the exact tokens. Each entry's RoPE position is its index among the candidates of its level, so at any length every relative distance stays within the range seen during training. Initialized from Qwen3-4B and trained on 0.62B tokens at the original model's 32K training length, RecSA outperforms all baselines on RULER at every length from 131K to 2M; at 1M, it scores 60.5, compared with 40.0 for the best baseline, Dense + DCA. With our optimized kernels, RecSA is faster than dense attention in end-to-end prefill and decoding from 64K to 2M; at 524K on one H200 GPU, it accelerates prefill by 3.2x and decoding by 3.8x.

cs.CL · cs.LG

新论文 · v1arXiv:2611.06458

Generative Learning on Dynamic Text-Attributed Graphs

Zhenyu Yang, Yuhong Chen, Binli Luo, Xiaodong Li, Xiaolin Hu, Hui Li, Jintang Li

原文摘要

Dynamic text-attributed graphs capture evolving interactions among entities with rich textual information, but future-link prediction remains difficult when nodes have little interaction history or appear as new targets. We propose \method, a generative framework that represents nodes with compact semantic identifiers (SIDs) learned from text and autoregressively predicts a target SID conditioned on the source SID, recent interaction events, and relative time gaps. A lightweight structural calibration further incorporates repetition, co-occurrence, and popularity. At inference time, \method encodes each query context once and reuses SID likelihoods to rank prescribed candidates, providing a candidate-efficient prediction interface. Experiments on nine datasets show broadly competitive performance across domains and graph densities. Source-side inductive evaluations demonstrate that text-derived SIDs provide a useful representation when a test source has no observed interaction history, while two recommendation-oriented case studies evaluate controlled new-user adaptation and first-arrival ranking for inductive new items. Further analyses characterize SID uniqueness, capacity, and temporal stability. Overall, the results support combining text-derived SIDs with temporal histories for dynamic link prediction.

cs.SI

新论文 · v1arXiv:2611.06381

LegalEvolver: Agentic Legal Judgment Prediction via Rubric-Guided Self-Evolution

Xinyu Yang, Jiejun Tan, Chenlong Deng, Chenghao Zhang, Tongyu Wen, Zhicheng Dou

原文摘要

As LLM-based agents develop, they have been increasingly applied in the legal domain. Legal Judgment Prediction~(LJP) requires models to conduct multi-step reasoning and external legal evidence retrieval. Existing legal search agents' performance is often constrained by the quality of retrieval queries. They often formulate queries directly from the initial user request, overlooking latent legal issues and legally salient details. To address this issue, we propose \emph{LegalEvolver}, a query optimization framework specifically designed to generate legally informative queries for LJP-oriented retrieval. Sampling high-quality optimized queries and their corresponding search trajectories is challenging, as relying solely on unguided high-temperature rollouts is both inefficient and constrained by the model's initial capabilities. We therefore heuristically optimize queries as seeds, improving the efficiency of trajectory generation. Furthermore, we summarize rollout experience into rubrics and use them to guide multi-round iterative self-evolution. Iteratively performing rubric optimization and rubric-guided RL enables the model to surpass its initial performance ceiling. End-to-end performance on extensive benchmarks demonstrates the effectiveness of query optimization on LJP. Comparisons with direct SFT followed by RL demonstrate the effectiveness of rubric guidance. Evolving rubrics further outperform fixed rubrics, confirming the benefits of self-evolution.

cs.AI · cs.CL