跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+479新增论文
+361其中 AI 相关
+64已有论文修订
+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 21:56各信源进度 →

统计口径与数据新鲜度

统计 2026-11-15 22:43 — 2026-11-16 22:43(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-16 01:34最近发现新增版本:2026-11-16 21:56库内论文最新更新时间:2026-11-14 02:59

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

23,827 已收录论文19,941 AI 相关6,922 有修订112 HF 社区发现

浏览论文库

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

找到 20 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)
新论文 · v1arXiv:2611.18139

PUMA-Former: Uncertainty-Aware Agentic Computation Offloading in MEO-LEO Satellite Networks With Incomplete Information

Yi Jing, Chunxiao Jiang, Jiawei Wang, Jiachen Sun

原文摘要

Satellite edge computing distributes mobile workloads across moving processors whose service capacity must be coordinated over time. In medium Earth orbit (MEO)--low Earth orbit (LEO) networks, periodic reports leave CPU states stale or missing, making offloading sensitive to uncertain capacity and shared-destination congestion. We propose PUMA-Former, an uncertainty-aware agentic framework for cooperative offloading under incomplete information that balances completed work, service delay, and execution cost. A MEO coordinator maintains distinct-report memory and closes the loop between available evidence, capacity estimates, allocation commands, and execution feedback. Latent-delay attention weights plausible measurement-age hypotheses, and historical innovations refine CPU estimates around the latest received report. Queue-error-aware fusion incorporates auxiliary pressure while accounting for backlog uncertainty. The inferred resource moments condition a structured Transformer allocator for joint destination, partition, bandwidth, CPU, and power decisions. Bounded risk-aware calibration then adjusts workload splits to reflect uncertainty and shared destination load. In traffic-driven development evaluation, PUMA-Former increases admitted compute-work completion by 8.62%, reduces mean modeled service delay by 34.78%, and reduces mean episode-P95 delay by 71.08% relative to ATLA-adapt. It also improves completion and normalized energy over fixed-half round-robin offloading. Service gains over the learning baselines involve higher normalized energy.

cs.NI · cs.AI

新论文 · v1arXiv:2611.17826

Space-based Zero-trust Architecture for Securing the Direct-to-Cell Paradigm

Xu Chen, Jiachen Sun, Linan Huang, Wei Quan, Yi Jing, Xuemin Shen

原文摘要

The rapid evolution of Direct-to-Cell (D2C) technology has emerged as a pivotal enabler for ubiquitous 6G global coverage, seamlessly connecting massive numbers of User Equipment (UEs) in remote and underserved areas. However, by bypassing traditional terrestrial satellite access gateways, the D2C paradigm inadvertently renders conventional front-end security perimeters obsolete. Consequently, malicious UEs can directly expose satellites and their core networks to severe security vulnerabilities. To address these critical threats, this article proposes a Space-based Zero-Trust Architecture (SZTA) empowered by on-board satellite computing. Specifically, the proposed SZTA consists of three core components: radio frequency (RF) fingerprint-based authentication, distributed identity management, and behavior-driven dynamic access control. This architecture enables continuous, real-time verification of UE identity authenticity and behavioral legitimacy, effectively mitigating the inherent network-layer security risks of D2C networks. Simulation results demonstrate the robustness, effectiveness, and efficiency of the proposed SZTA. Finally, to facilitate the practical deployment of {the SZTA}, we discuss key technical challenges, propose corresponding solutions, and outline open research issues to guide future investigations.

cs.NI

版本更新 · v3arXiv:2611.03233

Axion 1.0: Open Agent Foundation Models, Simplified

Yuqi Jia, Yunjie Tian, Jiayu Zheng, Yijun Liang, Jingchen Sun, Mengyu Zhao, Di Fu

原文摘要

We introduce Axion 1.0, a fully open agent foundation model that combines strong and broad agentic capabilities with a remarkably simple training recipe. Axion 1.0 is trained in only three stages: two lightweight supervised fine-tuning stages followed by reinforcement learning, without dedicated mid-training, direct preference optimization (DPO), or on-policy distillation (OPD). The first SFT stage develops strong self-contained reasoning through model-aware data construction, where we synthesize challenging problems and adversarially strengthen them based on the base model's own successful solution strategies. The second stage extends these capabilities to tool-dependent reasoning through a scalable pipeline for constructing diverse and automatically verifiable tasks grounded in real-world data sources and deterministic computation. Finally, reinforcement learning further strengthens long-horizon reasoning and interaction with external tools and environments. Despite this compact training recipe, Axion 1.0 achieves strong performance across a broad suite of agentic benchmarks and is competitive with leading commercial models on several of them. Equally importantly, Axion 1.0 is open beyond model weights: we release the complete recipe, including data construction, generation, filtering, verification, prompts, and training code, enabling the entire development process to be reproduced and extended. Axion 1.0 demonstrates that strong agent foundation models can be both open and simple, offering a practical and reproducible path toward advancing open agentic intelligence.

cs.AI

新论文 · v1arXiv:2611.14136

HaiRuo: Forecasting Ocean Eddies as Fields and Objects with a Satellite-Altimetry Foundation Model

Xukun Qin, Haochen Sun, Yuyang Tian, Ge Chen

原文摘要

Mesoscale eddies carry most of the ocean's kinetic energy and move heat, salt and nutrients across basins, yet where an eddy will be two weeks ahead is still read off ocean models with post-hoc tracking heuristics. We introduce HaiRuo, which forecasts eddies as fields and as objects at once: from 30 days of satellite sea-level maps, one forward pass returns 15 days of sea level and geostrophic velocity, daily eddy masks and the trajectory of every eddy alive today. HaiRuo pairs two components with a clear division of labour: (1) HaiRuo-1B, an open-weight 1.05-billion-parameter foundation model pretrained without labels on 24 years of global satellite-altimetry observations, which supplies the 15-day physical fields and masks (+0.14 field skill and +1.1 mIoU points under the same decoder) and further lowers 15-day centre error by 1.2 km frozen (1.35 km over training from scratch); and (2) XuanJi, a Lagrangian object decoder that samples the evolving forecast in each eddy's co-moving polar frame and refines the 15-day path across three look-ahead rounds, removing 2.5 km on HaiRuo-1B (2.9 km with joint fine-tuning) and 2.5-3.3 km via its object attention across six published backbones. Under one pre-registered global evaluation over the whole ocean, HaiRuo improves on fourteen methods on every endpoint (25.2 vs. 30.6 km 15-day centre error, -18%), holding its margin outside the labelled regions and through both phases of the El Nino-Southern Oscillation (ENSO), while reproducing westward drift, 50-60 deg S eastward reversal and poleward cyclone-anticyclone deflection. Over ten days through one fixed tracker, HaiRuo's forecast field cuts the eddy-centre error of WenHai and XiHe by 37-43% (-2.8 km vs. day 0 while they degrade +4.3-8.7 km), and its object trajectories cut pipeline error by 77-78% globally. Open-weight checkpoints, code and project page: https://github.com/xkqin/HaiRuo

cs.LG · cs.CV

新论文 · v1arXiv:2611.14042

Resilient Integrated Access and Backhaul Network Planning via Graph Neural Network Curriculum Learning

Jie Zhang, Haochen Sun, Mingzhang Zhou, Paul D. Mitchell, Hamed Ahmadi, Qiang Ni

原文摘要

Integrated Access and Backhaul (IAB) is a key enabler for cost-effective millimeter-wave (mmWave) densification, yet practical deployments are strongly constrained by the spatial availability of fiber-connected IAB donors. Variations in donor topologies reshape both the initial coverage footprint and the feasible multi-hop backhaul expansion patterns, causing planning policies trained on a single topology to generalize poorly. This paper studies cross-topology generalization for resilience-aware IAB deployment under coverage and transport feasibility requirements. We formulate IAB deployment as a constrained Markov decision process that couples access coverage satisfaction with multi-hop transport feasibility, including hop-by-hop rate-budget degradation and a minimum parent-connectivity requirement for fault tolerance. To solve the resulting large discrete planning problem, we propose a graph-based actor-critic approach that uses a Graph Attention Network v2 (GATv2) encoder with Proximal Policy Optimization (PPO), and we develop a multi-stage curriculum learning strategy that progressively increases constraint difficulty and scenario diversity for stable training and transferable policies. In a 60 GHz urban simulation over a 1000 m x 1000 m area with multiple donor layouts, the trained policy satisfies the minimum coverage requirement of 98% with fewer deployed IAB nodes than learning and heuristic baselines, and maintains higher coverage under progressive backhaul failures. The results demonstrate that curriculum-trained graph-based reinforcement learning can improve deployment efficiency and resilience while transferring across heterogeneous IAB donor topologies without retraining.

cs.NI

新论文 · v1arXiv:2611.11208

Multi-Granularity Communication-Computing Resource Mapping for User Association in MEO-Assisted LEO Satellite Networks

Yi Jing, Chunxiao Jiang, Jiawei Wang, Jiachen Sun, Yafeng Zhan

原文摘要

Low-Earth-orbit (LEO) satellite networks are evolving into spaceborne communication-computing infrastructures. Their user association must jointly account for the radio link and the CPU, GPU, NPU, memory, cache, storage, energy, queue, and thermal states that determine executable service capacity. This paper studies a MEO-assisted LEO architecture in which medium-Earth-orbit (MEO) regional state aggregators collect and relay time-stamped telemetry, the network operation control center (NOCC) maintains the global logical resource view and makes the final association decisions, and LEO satellites provide access and onboard execution. We develop a multigranularity representation and service-primitive-driven mapping that converts a service-level request into a target-LEO-dependent quantized resource vector. The integer formulation, exact mixed-integer linear programming (MILP) reference, and a capacity-aware primal-dual algorithm share the same mapped objective. Experiments use real-world service bandwidth traces, a 1584-satellite LEO shell, multiple traffic seeds, controlled load and mapping error, delayed telemetry, and temporal sessions. Relative to static mapping, the proposed method improves realized utility by 10.64% and lowers end-to-end outage by 5.70 percentage points. Target quantization reduces signaling by 19.84% with negligible utility loss. Additional load, delay, temporal, runtime, and exact-MILP audits delimit the operating region and remaining mapping uncertainty.

cs.NI

版本更新 · v2arXiv:2611.07482

SAVeR$^2$: Reasoning-based Safety Alignment for Large Reasoning Models via Verifiable Rewards

Yichen Sun, Qirui Liu, Yan Wang, Jianan Lin, Linbo Jiang, Shiyu Wang, Zhibo Wang, Zhixuan Chu

原文摘要

Large Reasoning Models (LRMs) have achieved impressive performance via explicit Chain-of-Thought (CoT) reasoning, yet this process introduces critical safety risks. We formalize LRM safety objectives as requiring non-malicious reasoning and answers, safety consistency, and explicit intent identification. We argue that existing alignment methods suffer from poor adversarial generalization, a significant ``safety tax'' on reasoning, and unfaithful coarse-grained rewards, primarily due to their neglect of the LRM's intrinsic reasoning capabilities for safety. To address these, we propose \textbf{SAVeR$^2$}, a novel \textbf{S}afety \textbf{A}lignment method via \textbf{R}ewards with \textbf{R}easoning capability. SAVeR$^2$ decomposes the safety objective into four fine-grained verifiable reward signals and utilizes a difficulty-aware data selection strategy to stabilize training. Crucially, we introduce a reasoning-preserving gradient projection mechanism that analytically resolves conflicts between safety and reasoning gradients, effectively eliminating the safety tax. Extensive experiments on models ranging from 1.5B to 671B demonstrate that SAVeR$^2$ achieves superior safety performance and significantly lower over-refusal rates across multiple benchmarks without compromising general reasoning capabilities.

cs.AI

版本更新 · v2arXiv:2608.14863

DDBench: A Benchmark for Agentic Debugging on Distributed Systems

Yibo Yan, Huijuan Wang, Junzhou He, Yizhuo Liang, Shaoyu Wang, Huanchen Sun, Seo Jin Park

原文摘要

LLM-based coding agents have advanced rapidly on single-process SWE tasks, with frontier models now clustering in the high-70s on SWE-bench Verified. Distributed-system debugging, however, remains an under-explored regime: bugs span processes, nodes, and protocol interactions, with root causes rarely recoverable from source alone and brute-force exploration intractable across non-deterministic interleavings. This leaves two gaps in LLM and agent evaluation: no code-repair benchmark targets distributed-system bugs, and no controlled study isolates how much externally provided debugging context changes agent success on them. We introduce DDBench, a code-repair benchmark of 60 historical bugs mined from 13 open-source distributed systems, partitioned into three reasoning-scope tiers. DDBench evaluates every case under two matched conditions: a symptom-only condition where the agent receives only the bug symptom and repository, and a context-augmented condition where it additionally receives bounded debugging context (logs, traces, runtime state, and targeted code-investigation notes), isolating the effect of debugging context from model capability. The evaluation of 10 LLMs on DDBench reveals several findings. First, distributed debugging exercises a reasoning dimension that single-process benchmarks do not surface: models' pass rates span 45 pp, and pairwise bootstrap separates 6 of 15 top-tier model pairs at p < 0.05 on DDBench's hardest case-set. Second, bounded debugging context lifts aggregate pass rate by +19.0 pp, and the lift is asymmetric: weaker models gain more in pass rate, while stronger models gain more in efficiency. Third, debugging context requires careful curation, as even faithful debugging context can anchor a strong model on the wrong layer of the system when its observation sits far from the fault.

cs.SE · cs.AI · cs.DC

新论文 · v1arXiv:2611.09340

Horizon-Coupled Obstacle Guidance for Relative-Action Diffusion Policies

Guangyuan Wang, Haishan Xia, Zitao Zhang, Kejia Chen, Yuanchen Sun, Zhenguo Sun, Donghao Zhang, Yu Zhang 等 9 位作者

原文摘要

Real-world robot deployment requires adapting learned policies to unseen obstacles without collecting new demonstrations or retraining. Existing policy-level guidance often applies obstacle corrections independently at individual horizon steps, which may address local obstacle violations but fail to coordinate the action sequence into a complete, task-directed detour. How obstacle responses should be propagated across the action horizon to coordinate complete, task-directed avoidance motion remains underexplored. We introduce an inferencetime obstacle-guidance framework for pretrained relative-action diffusion policies. It rolls out cumulative relative actions into a workspace trajectory, evaluates obstacle interactions along the trajectory with an SDF-based obstacle model, and propagates each obstacle response to all contributing action increments during denoising. Real-robot experiments across seven obstacleconstrained scenarios spanning four manipulation tasks show that our method achieves 105/140 (75.0%) collision-free task successes, compared with 8/140 (5.7%) for absolute-action pointwise guidance. Ablations show that the relative-action horizoncoupled formulation provides the larger observed improvement across both obstacle models.

cs.RO

版本更新 · v2arXiv:2610.14629

Executable but Unfaithful: Mock Object, Stub, and Assertion Generation by LLMs and Coding Agents

Hao Ning, Jiahui Dong, Zhen Yang, Guanda Yan, Yichen Sun, Xuanyu Zhu, Zheng Shen, Yifei Pei 等 10 位作者

原文摘要

Mocking is a foundational practice in unit testing, enabling developers to isolate the code under test by replacing real dependencies with simulated alternatives. Automating the generation of mock objects, stubs, and mock assertions has been a long-standing research goal. However, capture-and-replay approaches rely on recording the runtime behavior of actual dependencies. Evolutionary approaches such as StubCoder overcome this requirement but constrain generation to predefined symbol pools. Large language models (LLMs) and coding agents offer an alternative by generating mocking artifacts from source code without requiring recorded dependency executions or fixed symbol pools. Yet, no systematic study has evaluated their capabilities across all three artifact types. To address this gap, we construct MockBench, a benchmark comprising 165 mock artifact generation tasks from real-world Java and Python projects, with validated reference mock objects, stubs, and mock assertions. We evaluate six LLMs and three coding agents in terms of executability, behavior fidelity, and interaction-fault detection capability, complemented by comparisons with existing tools and sensitivity analysis. Experimental results show that LLMs and coding agents indeed have been the state-of-the-art techniques in mock-based testing. More specifically, compared with LLMs, coding agents substantially improve executability, while these gains do not consistently translate into higher behavior fidelity or better interaction-fault detection. Furthermore, we analyze 660 failures from LLMs and coding agents, and identify ten error types across the three artifact categories, revealing behavioral deviations that passing tests can conceal. Our findings highlight the challenges of mock artifact generation and provide practical guidance for both developers and researchers.

cs.SE

版本更新 · v2arXiv:2610.39550

Beyond Preference Following: Benchmarking Profile-Query Arbitration in Personalized Text Transformation

Tenghao Ji, Qichen Sun, Zhuoran Lu

原文摘要

Writing assistants use persistent preferences to personalize responses, but these can conflict with the current request. Responses can meet every requirement yet serve the task poorly, while task failures need not involve following a conflicting preference. We introduce Profile Conflict Arbitration (PCA-Bench), a benchmark of 64 rewriting tasks and 256 task-preference pairs. We construct these pairs by varying the profile preference for each fixed text and request, covering different relations between the preference and the request. We assess whether responses meet requirements, how well they achieve the request's goals, and whether they follow preferences. Across three LLMs, we find that when preferences compete with task goals, a higher success rate in meeting both requirements and preferences does not mean better responses: it can come with a greater quality loss compared with responses without a profile. Similar rates of preference following do not imply the same ability to meet the current task's requirements when preferences conflict with them.

cs.HC

新论文 · v1arXiv:2611.02761

Movable Antenna-Assisted Secret Key Generation over Temporally Correlated Channels

Zhiyu Huang, Chen Sun, Guyue Li, Guanghui Wen, Xianbin Wang

原文摘要

Physical-layer key generation (PLKG) exploits reciprocal channel dynamics to generate symmetric keys, addressing the key distribution problem in wireless communications. However, strong temporal correlation in slow-fading environments limits fresh randomness and results in low key generation rates. Movable antennas (MAs) introduce spatial degrees of freedom through antenna repositioning, providing a new approach to addressing the limitation. This paper investigates MA-assisted PLKG over temporally correlated channels. A first-order Gauss-Markov process is utilized to characterize the temporal correlation of multipath gains. We characterize the full-sequence mutual information (MI) over temporally correlated channels. For fixed antenna positions and beamforming, we establish a finite-horizon approximation error bound that vanishes as both effective signal-to-noise ratios (SNRs) increase. We derive a closed-form expression for the one-step conditional MI and adopt it as a tractable key-information metric. A nonconvex problem is formulated to jointly optimize antenna positions and per-slot beamforming under transmit-power and minimum-spacing constraints. We develop an alternating optimization algorithm, combining an exact principal-eigenvector beamforming solution with feasible SCA-based projected gradient position updates. Simulation results demonstrate that the MA-assisted scheme outperforms the fixed-position antenna baseline. At a high temporal correlation coefficient, position optimization achieves approximately $51\%$ higher rates than fixed antennas. Joint position and beamforming optimization provides a further $106\%$ gain over position optimization alone, demonstrating more pronounced gains under strong temporal correlation.

cs.IT · cs.CR · eess.SP

版本更新 · v3arXiv:2610.31028

Multimodal Progress-Residual Fusion for Anticipating Navigation Retreat in Robotic Flexible Endoscopy

Chi Kit Ng, Tianchun Wu, Jiachen Sun, Yuming Sun, Jianru Zhang, Hongliang Ren

原文摘要

Flexible robotic endoscopes provide access through tortuous anatomy, where friction, deformation, and wall interaction separate commanded insertion from useful distal progress. Operators interpret this mismatch and retreat to recover a navigable configuration. Recognizing the approach to that decision requires relating motor intent to distal response and the local visual scene. We introduce Multimodal Progress-Residual Fusion (MPRF), a causal model that anticipates operator retreat from commands, electromagnetic (EM) tracking, and endoscopic video. A nominal-progress model predicts the response expected in normal or successful navigation; a temporal predictor combines its EM residual with all three observed modalities. In a 150-episode colon-phantom collection, MPRF achieves 0.580 +/- 0.007 AUPRC across three seeds, compared with 0.511 for the strongest reduced-modality control. Complete multimodal histories provide the principal ranking advantage, while the residual adds an explicit intent-response representation with a small, uncertain increment over raw fusion. Analysis of frozen alarms relates warning timing to the behavioral label and explains the increase in event recall under a longer matching window. These results establish a basis for behavioral assistance during operator-controlled navigation.

cs.RO

新论文 · v1arXiv:2610.38391

SCOPE: Selective Critical-Step On-Policy Self-Distillation for Agentic Reinforcement Learning

Qirui Liu, Wei Cao, Yu Mi, Yichen Sun, Xuekun Zhang, Lei Liu, Yue Shen, Zhixuan Chu

原文摘要

Outcome-based reinforcement learning provides coarse supervision for long-horizon LLM agents: failed trajectories may contain correct steps, while successful trajectories can contain errors. Trajectory-level skills can also shift teacher preferences at unrelated steps, introducing off-target distillation signals. We introduce SCOPE, a framework for Selective Critical-Step On-Policy Self-Distillation. After a supervised warm-up, the policy analyzes its own trajectories, generates step-specific hindsight skills, and distills their guidance only at selected error steps. A bounded signed teacher-student coefficient stabilizes auxiliary updates while preserving positive reinforcement and negative correction. A success-relative pairwise length reward further encourages efficient task completion. Experiments on ALFWorld and WebShop demonstrate consistent improvements across multiple model backbones. With Qwen2.5-3B, SCOPE achieves 97.7% success on ALFWorld and 85.2% on WebShop, outperforming SEED by 5.9 and 6.3 percentage points, respectively. On unseen ALFWorld tasks, SCOPE reaches 89.6% success, improving over GRPO by 23.2 percentage points with 46.7% fewer interactions. Scaling to Qwen3-32B, SCOPE achieves 99.2% success on seen ALFWorld tasks and 94.0% on unseen tasks.

cs.AI

新论文 · v1arXiv:2610.36837

ExecDiffuser: Closing the Sample-to-Execution Gap in Diffusion Planning

Shuo Pei, Bosun Liang, Saihui Zha, Chen Sun, Huachun Tan, Yong Wang

原文摘要

Generative planners are often evaluated by the trajectories they can generate, although closed-loop driving ultimately depends on the single candidate committed to control. We identify this mismatch as the sample-to-execution gap and show that improving proposal quality alone can worsen the executed decision. Here, we introduce ExecDiffuser, a recurrent proposal-to-execution framework that couples candidate generation, trajectory selection, and subsequent replanning through shared execution semantics. Intent-Guided Diffusion Proposal (IGD) assigns hypothesis responsibility and shapes trajectories with execution-conditioned supervision. Execution-Grounded Selection (EGS) learns a candidate-conditioned publication utility, commits one trajectory to control, and returns it as context for the next planning cycle. Across nuPlan, InterPlan, and NAVSIM, ExecDiffuser achieves leading closed-loop performance against strong state-of-the-art baselines. Under a unified five-seed protocol, it obtains an average score of 85.62, improving over Diffusion Planner and FlowDrive by 4.02 and 1.31 points, respectively. On NAVSIM, it achieves 92.6 PDMS under the official evaluation pipeline. Matched counterfactual analysis further shows that ExecDiffuser more reliably converts strong candidates into committed decisions, while prefix interventions confirm that this advantage persists after physical execution and subsequent replanning.

cs.RO

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1106ACL Anthology · ACL 2026

AT²PO: Agentic Turn-based Policy Optimization via Tree Search

Zefang Zong Tencent Inc., Dingwei Chen SUN YAT-SEN UNIVERSITY, Yang Li Tencent Inc., Qi Yi University of Science and Technology of China, Bo Zhou, Chengming Li Shenzhen MSU-BIT University, BO Qian Tencent Inc., Peng Chen 等 9 位作者

原文摘要

LLM agents have emerged as powerful systems for tackling multi-turn tasks by interleaving internal reasoning and external tool interactions. Agentic Reinforcement Learning has recently drawn significant research attention as a critical post-training paradigm to further refine these capabilities. In this paper, we present AT²PO (Agentic Turn-based Policy Optimization via Tree Search), a unified framework for multi-turn agentic RL that addresses three core challenges: limited exploration diversity, sparse credit assignment, and misaligned policy optimization. AT²PO introduces a turn-level tree structure that jointly enables Entropy-Guided Tree Expansion for strategic exploration and Turn-wise Credit Assignment for fine-grained reward propagation from sparse outcomes. Complementing this, we propose Agentic Turn-based Policy Optimization, a turn-level learning objective that aligns policy updates with the natural decision granularity of agentic interactions. ATPO is orthogonal to tree search and can be readily integrated into any multi-turn RL pipeline. Experiments across seven benchmarks demonstrate consistent improvements over the state-of-the-art baseline by up to 1.84 percentage points in average, with ablation studies validating the effectiveness of each component.

cs.CL · cs.AI

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.190ACL Anthology · ACL 2026

Stable Language Guidance for Vision–Language–Action Models

Zhihao Zhan, Yuhao Chen SUN YAT-SEN UNIVERSITY, Jiaying Zhou SUN YAT-SEN UNIVERSITY, Qinhan Lyu, Hao Liu SUN YAT-SEN UNIVERSITY, Keze Wang SUN YAT-SEN UNIVERSITY, Liang Lin SUN YAT-SEN UNIVERSITY, Guangrun Wang SUN YAT-SEN UNIVERSITY

原文摘要

Vision-Language-Action (VLA) models have demonstrated impressive capabilities in generalized robotic control; however, they remain notoriously brittle to linguistic perturbations. We identify a critical "modality collapse” phenomenon where strong visual priors overwhelm sparse linguistic signals, causing agents to overfit to specific instruction phrasings while ignoring the underlying semantic intent. To address this, we propose Residual Semantic Steering (RSS), a probabilistic framework that disentangles physical affordance from semantic execution. RSS introduces two theoretical innovations: (1) Monte Carlo Syntactic Integration, which approximates the true semantic posterior via dense, LLM-driven distributional expansion, and (2) Residual Affordance Steering, a dual-stream decoding mechanism that explicitly isolates the causal influence of language by subtracting the visual affordance prior. Theoretical analysis suggests that RSS effectively maximizes the mutual information between action and intent while suppressing visual distractors. Empirical results across diverse manipulation benchmarks demonstrate that RSS achieves state-of-the-art robustness, maintaining performance even under adversarial linguistic perturbations.

cs.CL · cs.CV

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.494ACL Anthology · ACL 2026

EchoVLM: Dynamic Mixture-of-Experts Vision-Language Model for Universal Ultrasound Intelligence

Chaoyin She Northwest Polytechnical University Xi'an, Ruifang Lu, Lida Chen SUN YAT-SEN UNIVERSITY, Wei Wang SUN YAT-SEN UNIVERSITY, Qinghua Huang Northwest Polytechnical University Xi'an

原文摘要

Ultrasound is the preferred early cancer screening modality due to non-ionizing radiation, cost-effectiveness, and real-time imaging, yet conventional diagnosis relies heavily on physician expertise, causing significant subjectivity and limited efficiency. Vision-Language Models (VLMs) show promise but lack ultrasound-specific knowledge and multi-organ generalization. We propose EchoVLM, the first open-source 10-billion-parameter ultrasound-tailored VLM with a Mixture-of-Experts (MoE) architecture. It is infused with knowledge across seven anatomical systems, trained on 208,941 clinical cases, 1.47 million ultrasound key-frame images, and over 100 diseases or imaging findings. Supporting clinical report generation, diagnosis prediction, and Visual Question Answering (VQA), it outperforms Qwen2-VL by 7.58 BLEU-1 and 3.45 ROUGE-1 points in report generation. This work shows substantial potential for establishing a general-purpose ultrasound VLM and lays a technical foundation for clinical translation. Source code and model weights are available at https://github.com/Asunatan/EchoVLM.

cs.CL · cs.CV

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.710ACL Anthology · ACL 2026

Act-Adaptive Margin: Dynamically Calibrating Reward Models for Subjective Ambiguity

Feiteng Fang, Dingwei Chen SUN YAT-SEN UNIVERSITY, Xiang Huang, Ting-En Lin, Yuchuan Wu Alibaba Group, Xiong Liu Alibaba Group, Jing Ye Institute of automation, Chinese academy of science, Chinese Academy of Sciences, Ziqiang Liu 等 13 位作者

原文摘要

Currently, most reinforcement learning tasks focus on domains like mathematics and programming, where verification is relatively straightforward. However, in subjective tasks such as role-playing, alignment techniques struggle to make progress, primarily because subjective reward modeling using the Bradley-Terry model faces significant challenges when dealing with ambiguous preferences. To improve reward modeling in subjective tasks, this paper proposes AAM (Act-Adaptive Margin), which enhances reward modeling by dynamically calibrating preference margins using the model’s internal parameter knowledge. We design two versions of AAM that efficiently generate contextually-appropriate preference gaps without additional human annotation. This approach fundamentally improves how reward models handle subjective rewards by better integrating generative understanding with preference scoring. To validate AAM’s effectiveness in subjective reward modeling, we conduct evaluations on RewardBench, JudgeBench, and challenging role-playing tasks. Results show that AAM significantly improves subjective reward modeling performance, enhancing Bradley-Terry reward models by 2.95% in general tasks and 4.85% in subjective role-playing tasks. Furthermore, reward models trained with AAM can help downstream alignment tasks achieve better results. Our test results show that applying rewards generated by AAM-Augmented RM to preference learning techniques (e.g., GRPO) achieves state-of-the-art results on CharacterEval and Charm. The code and dataset will be released upon acceptance.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.911ACL Anthology · ACL 2026

IS-CoT: Breaking the Long-form Generation Collapse via Interleaved Structural Thinking

Zechen Sun, Yuyang Sun, Zecheng Tang Soochow University, Juntao Li, Wenpeng Hu Academy of Military Science, Wenliang Chen Soochow University, China, Zhunchen Luo, Guotong Geng AMS 等 9 位作者

原文摘要

Generating coherent and controllable long-form content remains a persistent challenge for Large Language Models (LLMs). While reasoning-enhanced models have demonstrated success in logic-intensive domains, our evaluation reveals that they suffer from a severe length collapse in open-ended writing, where performance degrades sharply as target lengths exceed 2,000 words. We attribute this failure to the limitation of static hierarchical planning, which struggles to provide dynamic guidance over extended contexts. To bridge this gap, we introduce the Interleaved Structural Chain-of-Thought (IS-CoT) framework. Unlike external agentic workflows, IS-CoT embeds a dynamic Plan-Write-Reflect cycle into the generation process, enabling continuous strategy adaptation and global alignment without additional assistance. Based on this framework, we construct a high-quality dataset of interleaved reasoning traces via a multi-teacher pipeline and train IS-Writer-8B. Experiments demonstrate that IS-Writer-8B achieves state-of-the-art performance on challenging long-form benchmarks (e.g., +3.08 vs. DeepSeek-V3.2 on LongBench-Write), exhibiting robust length compliance and coherence competitive with significantly larger proprietary models.

cs.CL · cs.AI

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表