探索全球前沿学术脉络

01.

arXiv (CS.CL) 2026-06-15 DOI: arXiv:2606.13931

DLawBench: Evaluating LLMs Through Multi-Turn Legal Consultation

作者:

Li Zhang↗Yuzhen Shi↗Yiran Hu↗Jingwen Zhang↗Wenbo Lv↗Yubo Ma↗Wei Wang↗Rongyao Shi↗Yuanyang Qiu↗Xinran Xu↗Yuemeng Qi↗Linlin Miao↗…

Lawyer-client consultation is a critical starting point for legal services. Effective legal assistance hinges on eliciting sufficient and truthful information from clients in order to devise strategies that best protect their interests. This task requires Large Language Models (LLMs) not only to perform robust legal reasoning, but also to strategically elicit material facts through multi-turn interactions and effectively guide clients with diverse personalities. Yet existing legal benchmarks overlook this interactive capability. To fill this gap, we introduce DLawBench, a diagnostic benchmark for real-world legal consultation. Drawing on realistic client behavior, we characterize lawyer-client interactions into four types: Cooperative, Dependent, Withdrawn, and Adversarial. Using dialogues grounded in real cases, DLawBench evaluates whether LLMs can effectively conduct legal consultation under realistic conditions. DLawBench comprises 461 cases from Chinese and U.S. law, 5,532 paired fact entries, 3,411 inquiry rubrics, and 3,348 issue-resolution rubrics, and evaluates 26 representative LLMs. Systematic experiments show substantial headroom: the best-performing model, GPT-5.5, achieves only 0.562 on consultation-grounded legal reasoning. More importantly, DLawBench exposes both sycophancy in legal consultation and a paradox: models perform worse when clients need guidance most.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2606.17386

TerraTransfer: Learning End-to-End Driving Policies Without Expert Demonstrations

作者:

Zikang Xiong↗Weixin Li↗Zhouchonghao Wu↗Akshay Rangesh↗Saarth Bonde↗Grantland Hall↗Chen Tang↗Yihan Hu↗Wei Zhan↗

End-to-end autonomous driving has achieved state-of-the-art performance on benchmarks and real-world deployments. Its standard training recipe, however, is expensive across all stages: collecting and labeling millions of driving frames is costly, and closed-loop RL on images is bottlenecked by the per-step cost of photorealistic rendering plus a forward pass through a large vision backbone. Self-play in vectorized simulators changes the economics: millions of rollout steps per second, and a state distribution naturally rich in collisions, near-misses, and recoveries that no driving log contains. Our approach exploits this asymmetry by decoupling learning to drive from learning to see. We pretrain a single policy by self-play, then align its latent space with a pretrained vision backbone, through the action KL divergence and a batch-relational low-rank structural loss. The action target comes from the self-play policy, so alignment never supervises against a logged trajectory: a paired dataset of (image, scene-state) frames suffices, with no need for the curated expert demonstrations that imitation pretraining is built on. On photorealistic 3D Gaussian splatting closed-loop scenarios, the resulting end-to-end policy matches or exceeds prior end-to-end methods.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.12087

FORT-Searcher: Synthesizing Shortcut-Resistant Search Tasks for Training Deep Search Agents

作者:

Jia Deng↗Yimeng Chen↗Xiaoqing Xiang↗Ziyang Zeng↗Shuo Tang↗Wayne Xin Zhao↗Feng Chang↗Chuan Hao↗Yuan Wei↗Ran Tao↗Bryan Dai↗Ji-Rong Wen↗…

Training deep search agents requires verifiable questions whose answers remain unavailable until sufficient evidence has been acquired through search. Existing synthesis methods often increase apparent difficulty by enriching graph structures, but structural complexity alone does not guarantee realized search difficulty: the intended search process can collapse through a cheaper identifying route. We formalize this gap with a shortcut-aware difficulty framework and identify four actionable shortcut risks: evidence co-coverage, single-clue selectivity, exposed constants, and prior-knowledge binding. To diagnose their realized effects, we use trajectory signatures including solving cost, answer hit time, and prior-shortcut rate. Guided by this framework, we introduce FORT, a Framework of Shortcut-Resistant Training-Data Synthesis. FORT constructs shortcut-resistant training data by controlling shortcut risks across entity selection, evidence graph construction, question formulation, and adversarial refinement. Experiments show that FORT induces longer pre-answer search and fewer shortcut patterns than existing open-source deep search datasets. Using the resulting trajectories, we train FORT-Searcher with supervised fine-tuning (SFT) only, and it achieves the best overall performance among comparable-size open-source search agents on challenging deep search benchmarks. Relevant resources will be made available at https://github.com/RUCAIBox/FORT-Searcher.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.CV) 2026-06-18 DOI: arXiv:2606.19053

Benchmarking Large Vision-Language Models on Fine-Grained Image Tasks: From Evaluation to Diagnosis

作者:

Hong-Tao Yu↗Chen-Wei Xie↗Yuxin Peng↗Serge Belongie↗Xiu-Shen Wei↗

Recent advancements in Large Vision-Language Models (LVLMs) have demonstrated remarkable multimodal perception and reasoning capabilities. While numerous benchmarks have evaluated LVLMs from holistic or task-specific perspectives, their capabilities on fine-grained image tasks-fundamental to computer vision-remain insufficiently understood. To address this gap, we introduce FG-BMK, a comprehensive fine-grained evaluation benchmark containing 1.01 million questions and 0.28 million images, covering diverse scenarios from common object-centric domains to specialized domains. FG-BMK jointly evaluates dialogue-level fine-grained semantic recognition and feature-level visual discriminability through human-oriented and machine-oriented paradigms, enabling diagnostic analysis of whether LVLM failures arise from insufficient visual representations, weak visual-to-semantic grounding, or limited fine-grained knowledge. Through extensive experiments on a diverse set of representative LVLMs/VLMs, we find that current LVLMs remain inadequate fine-grained recognizers, with failures arising from intertwined bottlenecks in visual representations, semantic grounding, modality alignment, and category-level knowledge. We further analyze training design factors for improving fine-grained capabilities and examine how visual and linguistic perturbations affect LVLM predictions. These findings provide diagnostic insights into the limitations of current LVLMs and offer guidance for future data construction and model design in developing more reliable LVLMs for fine-grained visual tasks. Our code is open-source and available at https://fg-bmk.github.io/.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.12370

Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling

作者:

Yucheng Li↗Huiqiang Jiang↗Yang Xu↗Jianxin Yang↗Yi Zhang↗Yizhong Cao↗Yuhao Shen↗Fan Zhou↗Rui Men↗Jianwei Zhang↗An Yang↗Bowen Yu↗…

Reinforcement learning (RL) has become a key component in modern large language models, yet the rollout stage remains the key bottleneck in RL training pipelines. Although Multi-Token Prediction (MTP) offers a natural solution to accelerate rollouts through speculative decoding, many studies have observed that MTP acceptance rates degrade significantly during RL training, leading to limited speedup performance. To address this bottleneck, we present Bebop, a systematic study of MTP in LLM post-training, and offer practical recipes to integrate MTP into large-scale RL pipelines. First, we reveal that the MTP acceptance rate is fundamentally bounded by the fluctuation of model entropy, which demonstrates a clear negative linear relationship with the rise of entropy in the RL stage. Second, we show that probabilistic rejection sampling largely alleviates the disturbance introduced by entropy in RL compared to greedy draft sampling. We further identify that the conventional MTP training objectives (cross-entropy or KL) are suboptimal in such settings, and therefore we propose a novel end-to-end TV loss that directly optimizes multi-step rejection sampling acceptance rate, yielding ~10% acceptance rate improvements, achieving up to 95% acceptance rates and up to 25% extra inference throughput gains across mathematical reasoning, code generation, and agentic tasks. Third, we test various online MTP training strategies during RL and show that pre-RL MTP training with e2e TV loss and rejection sampling achieves a consistent acceptance rate and speedup throughout the entire RL, eliminating the need for costly online MTP updating. We provide extensive experiments and analysis that validate our findings. Experimental results show our method achieves up to 1.8x end-to-end acceleration in async RL training of Qwen3.5, Qwen3.6, and Qwen3.7 models.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.15652

MosaicQuant: Inlier-Outlier Disaggregation for Unified 4-Bit LLM Quantization

作者:

Yangjia Hu↗Haodong Wang↗Zicong Hong↗Qianli Liu↗Quanxin Shou↗Jian Lin↗Song Guo↗Xiaowei Shen↗Xiangjun Huang↗Dian Wang↗Jian Yang↗

4-bit quantization significantly reduces the memory footprint and accelerates the inference of large language models (LLMs). However, its limited bit-width representation struggles to faithfully capture both dense common values (inliers) and rare large-magnitude values (outliers), causing substantial accuracy degradation. Existing mixed-precision methods mitigate this by retaining outliers in high precision, but at the cost of breaking the uniformity of low-bit execution, introducing precision conversion and extra data movement that undermine practical speedup. We propose MosaicQuant, a unified 4-bit LLM quantization paradigm built on a novel principle of inlier–outlier disaggregation. Rather than elevating outlier precision, MosaicQuant quantizes the full weight matrix into a dense 4-bit base component, where inliers are captured faithfully while outlier are inevitably quantized. A sparse 4-bit residual component is then introduced to compensate for these quantization errors, selectively targeting the most error-critical weight blocks where output distortion is shown to be concentrated. However, a unified representation alone is insufficient, as naïvely executing the sparse residual as a separate kernel still breaks the unified low-bit inference pipeline. To bridge this gap, we introduce ZipperEngine, which fuses sparse block computation into the dense 4-bit GEMM kernel via an overlapped pipeline, unifying not only the representation but also the execution into a single coherent low-bit inference pipeline. Extensive experiments on LLaMA3 and Qwen3 demonstrate that MosaicQuant preserves near-FP16 accuracy while achieving up to $1.24\times$ speedup over the W16A16 baseline.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2509.11575

A Survey of Reasoning and Agentic Systems in Time Series with Large Language Models

作者:

Ching Chang↗Yidan Shi↗Defu Cao↗Wei Yang↗Jeehyun Hwang↗Haixin Wang↗Jiacheng Pang↗Wei Wang↗Yan Liu↗Wen-Chih Peng↗Tien-Fu Chen↗

arXiv:2509.11575v3 Announce Type: replace Abstract: Time series reasoning treats time as a first-class axis and incorporates intermediate evidence directly into the answer. This survey defines the problem and organizes the literature by reasoning topology with three families: direct reasoning in one step, linear chain reasoning with explicit intermediates, and branch-structured reasoning that explores, revises, and aggregates. The topology is crossed with the main objectives of the field, including traditional time series analysis, explanation and understanding, causal inference and decision making, and time series generation, while a compact tag set spans these axes and captures decomposition and verification, ensembling, tool use, knowledge access, multimodality, agent loops, and LLM alignment regimes. Methods and systems are reviewed across domains, showing what each topology enables and where it breaks down in faithfulness or robustness, along with curated datasets, benchmarks, and resources that support study and deployment (https://github.com/blacksnail789521/Time-Series-Reasoning-Survey). Evaluation practices that keep evidence visible and temporally aligned are highlighted, and guidance is distilled on matching topology to uncertainty, grounding with observable artifacts, planning for shift and streaming, and treating cost and latency as design budgets. We emphasize that reasoning structures must balance capacity for grounding and self-correction against computational cost and reproducibility, while future progress will likely depend on benchmarks that tie reasoning quality to utility and on closed-loop testbeds that trade off cost and risk under shift-aware, streaming, and long-horizon settings. Taken together, these directions mark a shift from narrow accuracy toward reliability at scale, enabling systems that not only analyze but also understand, explain, and act on dynamic worlds with traceable evidence and credible outcomes.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.15129

EyeMVP: OCT-Informed Fundus Representation Learning via Paired CFP–OCT Pretraining

作者:

Zhuo Deng↗Ruiheng Zhang↗Ziheng Zhang↗Weihao Gao↗Yitong Li↗Qian Wang↗Lei Shao↗Jiaoyue Dong↗Zhixi Zeng↗Lijian Fang↗Haibo Wang↗Xiaobin Lin↗…

Color fundus photography (CFP) is the mainstay for large-scale retinal screening, yet its diagnostic capacity is constrained by the lack of depth-resolved structural information. Optical coherence tomography (OCT) provides cross-sectional retinal anatomy, but is less accessible in population-level screening. Here, we present EyeMVP, a cross-modal retinal foundation model that uses paired CFP–OCT pretraining to learn OCT-informed CFP representations. EyeMVP is pretrained on 674,893 strict same-eye same-day paired CFP–OCT image triples from 112,642 patients across eight hospitals in China. The model uses cross-modal masked reconstruction to enrich CFP representations with OCT-associated supervision, while requiring only CFP images at inference. To accommodate the non-aligned imaging geometry between en-face CFP and cross-sectional OCT, EyeMVP combines source-constrained cross-attention with CFP-derived structural masks. Across 16 downstream tasks, including classification, segmentation, few-shot adaptation, and cross-modal retrieval, EyeMVP outperforms representative retinal foundation models and shows consistent gains on tasks involving macular and optic nerve structure. For CFP-challenging macular diseases, EyeMVP achieves an AUROC of 0.948 for macular edema (vs.~0.852 for EyeCLIP) and 0.825 for myopic macular schisis. In an exploratory reader study, EyeMVP exceeds junior and intermediate ophthalmologist groups but does not reach senior ophthalmologist performance on macular edema, while showing numerically higher balanced accuracy than all reader groups on myopic macular schisis. These results suggest that pixel-level cross-modal reconstruction can enrich CFP representations with OCT-associated supervision, providing a practical route toward stronger CFP-based retinal analysis in screening settings.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2508.00956

FOUNDv2: Learning Unified User Quantized Tokenizers for User Representation

作者:

Chuan He↗Yang Chen↗Bin Dou↗Wuliang Huang↗Baokun Wang↗Yongchao Liu↗Xing Fu↗Yu Cheng↗Chuntao Hong↗Weiqiang Wang↗Zhongle Xie↗Jiajun Zheng↗…

arXiv:2508.00956v3 Announce Type: replace-cross Abstract: User representation learning serves as a fundamental pillar for personalized services on large-scale web platforms. Despite its importance, conventional continuous embedding methods face significant challenges, including the lack of a unified paradigm for multi-source data integration, prohibitive storage overhead due to low information density, and the lack of multi-scale modeling granularity. To overcome these limitations, we introduce FOUNDv2, a comprehensive user representation scheme centered on the Unified User Quantized Tokenizer U2QT) framework. FOUNDv2 transforms heterogeneous user data into a standardized discrete token space through a robust two-stage architecture. Specifically, the framework first extracts compact feature representations and subsequently employs a multi-view RQ-VAE to discretize them into storage-efficient tokens using shared and source-specific codebooks. To empower these representations with predictive intelligence, we further design multi-scale alignment objectives to capture both fine-grained behavioral dependencies and macro-temporal periodicity. Extensive experiments on various benchmarks demonstrate that FOUNDv2 consistently outperforms task-specific baselines while achieving substantial reductions in storage and computational costs. Finally, the large-scale deployment of FOUNDv2 on Alipay validates its practical scalability and efficiency across diverse industrial scenarios. The main code is available at: https://github.com/chuanhe1999/FOUNDv2.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.04075

Large Language Models Hack Rewards, and Society

作者:

Wei Liu↗Xinyi Mou↗Hanqi Yan↗Zhongyu Wei↗Yulan He↗

Reinforcement learning (RL) has become a dominant post-training paradigm, enabling large language models (LLMs) to learn from rewards. We observe that societal regulations are structurally similar to reward functions. They define measurable outcomes, thresholds, and exceptions, while often leaving institutional intent only partially specified. We hypothesise that the RL training process may exploit these gaps and therefore ask whether models' well-known tendency to hack reward functions during RL can scale into a more consequential failure mode named societal hacking: discovering loopholes in the rules society runs on. To study this phenomenon, we introduce SocioHack, a sandbox of 72 societal environments, and find that within these environments, reward hacking naturally emerges and leads to regulatory loophole discovery. Models learn to hack the social rules and generate strategies that remain technically compliant while defeating regulatory intent, and current LLM safeguards provide only limited mitigation. Therefore, collecting in-the-wild feedback for model training requires greater caution, and we need a next-generation post-training paradigm for safely iterating LLMs in real society.=

阅读与讨论 → 访问原文 →

11.

arXiv (CS.CV) 2026-06-15 DOI: arXiv:2606.14251

HiST: A Hierarchical Sparse Transformer for Cross-Modal Spatial Transcriptomics Modeling

作者:

Weiyi Wu↗Xinwen Xu↗Xingjian Diao↗Siting Li↗Zhi Wei↗Alma Andersson↗Jiang Gui↗

Spatial transcriptomics (ST) links gene expression with tissue morphology but remains expensive and low-throughput, motivating surrogates that infer expression from routine histology. Whole-slide H&E-to-ST inference pairs a gigapixel image with gene measurements at a sparse, irregular set of locations, making multiscale modeling challenging without incurring dense-grid overhead or quadratic token mixing. We propose HiST, a hierarchical sparse transformer that treats measured locations as a lattice-indexed sparse field and builds a dyadic encoder–decoder directly on the active tissue footprint. HiST combines sparse window attention for local geometric correspondence with resolution-changing operators for rapid multiscale context integration. For a fixed window size, the dominant runtime and memory scale with the number of observed locations rather than the dense slide area. To mitigate slide-specific acquisition variation, HiST adds a bottlenecked global conditioning pathway via a slide calibration token that summarizes slide-level context and conditions local representations. On a multi-organ benchmark spanning diverse tissues and acquisition sources, HiST improves predictive performance over recent baselines while reducing runtime and peak memory.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2606.12942

PRISMR: Overcoming Parse Collapse in Multimodal Listwise Ranking via Parameterized Representation Internalization

作者:

Hao Jiang↗Xin Li↗Annan Wang↗Zhi Yang↗Haoxiang Zhang↗Yichi Zhang↗Weisi Lin↗

arXiv:2606.12942v1 Announce Type: new Abstract: Generative listwise ranking with Large Multimodal Models (LMMs) aims to capture global list context in a single forward pass, but its effectiveness degrades in long-context multimodal scenarios. We identify a recurring failure mode, parse collapse, where the autoregressive decoder produces fluent yet incomplete rankings by silently omitting candidates and terminating early. This failure stems from limited context utilization rather than simple formatting mistakes, making prompt engineering and constrained decoding insufficient. We propose PRISMR (Parameterized Representation Internalization for Semantic Multimodal Ranking), a framework that replaces transient in-context list processing with parametric structural conditioning. PRISMR uses a lightweight hypernetwork to encode multimodal candidates in parallel and generate item-specific LoRA weights, which are synthesized into an instance-specific adapter for a LMM. This paradigm enables more robust internalization of list structure while preserving the base model. We further introduce a large-scale multimodal review-ranking benchmark for evaluation. Experiments demonstrate that PRISMR substantially reduces parse collapse, improves listwise ranking performance, and transfers effectively across domains and instruction-tuned backbones.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.CV) 2026-06-15 DOI: arXiv:2606.14383

IndustryBench-MIPU: Benchmarking Multi-Image Attribute Value Extraction for Industrial Products

作者:

Haonan Qi↗Jin Cao↗Yongqi Zhang↗Xintong Wang↗Weidong Tang↗Bin Chen↗Chengfu Huo↗Haojun Pan↗Hengyu You↗Jing Li↗Yingde Wang↗Liang Ding↗…

Industrial products such as valves and circuit breakers are defined by dense technical specifications that govern procurement, compatibility, and safety across supply chains. These specifications are scattered across multiple heterogeneous product images, including specification tables, nameplates, and technical drawings, yet whether Multimodal Large Language Models (MLLMs) can reliably recover them remains underexplored. To fill this gap, we introduce IndustryBench-MIPU, the first large-scale benchmark for multi-image industrial product understanding, built around structured attribute extraction – recovering property-value pairs from product images. This task jointly probes text recognition on specification tables and nameplates, visual reasoning over technical drawings, domain knowledge to decode industrial terminology, and cross-image evidence integration to assemble scattered specifications. Concretely, the benchmark comprises 4,559 products across 27,652 images with 103,703 annotations spanning 18 industrial categories, constructed through multi-model consensus and three-tier quality assurance. Evaluating nine MLLMs under both single-image and product-level multi-image settings reveals a stark completeness gap: models achieve high precision (86–94%) but the best recovers only 49.9% of product-level attributes; moving from single-image to multi-image extraction costs 15–34 percentage points of recall. Multi-image completeness, not single-image accuracy, is the core bottleneck. Dataset and code are publicly available.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2602.18746

Bridging Modality Disconnect in Self-Reflection via Closed-Loop Visually Grounded Verification

作者:

Haoyu Zhang↗Yuwei Wu↗Pengxiang Li↗Xintong Zhang↗Zhi Gao↗Rui Gao↗Mingyang Gao↗Che Sun↗Yunde Jia↗

In the era of Vision-Language Models (VLMs), enhancing multimodal reasoning capabilities remains a critical challenge, particularly in handling ambiguous or complex visual inputs, where initial inferences often lead to hallucinations or logic errors. Existing VLMs often produce plausible yet ungrounded answers, and even when prompted to "reflect", their corrections may remain detached from the image evidence. To address this, we propose the MIRROR framework for Multimodal Iterative Reasoning via Reflection On visual Regions. By embedding visual reflection as a core mechanism, MIRROR is formulated as a closed-loop process comprising draft, critique, region-based verification, and revision, which are repeated until the output is visually grounded. To facilitate training of this model, we construct **ReflectV**, a visual reflective dataset for multi-turn supervision that explicitly contains reflection triggers, region-based verification actions, and answer revision grounded in visual evidence. Experiments on both general vision-language benchmarks and representative vision-language reasoning benchmarks show that MIRROR improves correctness and reduces visual hallucinations, demonstrating the value of training reflection as an evidence-seeking, region-aware verification process rather than a purely textual revision step.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2511.20709

DualGauge: Automated Joint Security-Functionality Benchmarking of Specification-Only Code Generation by LLMs and Coding Agents

作者:

Rupam Patir↗Keyan Guo↗Suvadra Barua↗Abhijeet Pathak↗Dinesh Gudimetla↗Jiawei Guo↗Hongxin Hu↗Haipeng Cai↗

arXiv:2511.20709v2 Announce Type: replace-cross Abstract: Large language models (LLMs) and LLM-based coding agents are now used to generate code from natural-language specifications, yet ensuring such code is both functionally correct and secure remains a challenge. We present DualGauge, the first fully automated framework for jointly evaluating correctness and security of specification-only code generation, supported by DualGauge-Bench, a language-agnostic benchmark of 307 coding tasks each paired with functional and security tests derived from the same specification. Evaluating 10 representative LLMs across Python, C++, and JavaScript, we find that functional correctness substantially overestimates reliable code generation: even the strongest model remains below 15% joint security-functionality success in every language. Common model-side factors–scale, extended thinking, quantization, instruction tuning, and code specialization–do not reliably improve joint performance, suggesting secure-and-correct code generation does not simply emerge from stronger coding capability. Evaluation of 3 leading agentic coding systems (Codex, OpenHands, and Claude Code) shows that iterative scaffolding provides no advantage over direct (LLM-based) generation on specification-only tasks. A qualitative audit reveals failures concentrate at the output contract boundary and in guards that exist but are insufficient–patterns that only joint benchmarking reliably exposes.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.LG) 2026-06-15 DOI: arXiv:2602.08324

Towards Efficient Large Language Reasoning Models via Extreme-Ratio Chain-of-Thought Compression

作者:

Yuntian Tang↗Bohan Jia↗Wenxuan Huang↗Lianyue Zhang↗Jiao Xie↗Wenxi Li↗Wei Li↗Jie Hu↗Xinghao Chen Rongrong Ji↗Shaohui Lin↗

arXiv:2602.08324v5 Announce Type: replace Abstract: Chain-of-Thought (CoT) reasoning successfully enhances the reasoning capabilities of Large Language Models (LLMs), yet it incurs substantial computational overhead for inference. Existing CoT compression methods often suffer from a critical loss of logical fidelity at high compression ratios, resulting in significant performance degradation. To achieve high-fidelity, fast reasoning, we propose a novel EXTreme-RAtio Chain-of-Thought Compression framework, termed Extra-CoT, which aggressively reduces the token budget while preserving answer accuracy. To generate reliable, high-fidelity supervision, we first train a dedicated semantically-preserved compressor on mathematical CoT data with fine-grained annotations. An LLM is then fine-tuned on these compressed pairs via a mixed-ratio supervised fine-tuning (SFT), teaching it to follow a spectrum of compression budgets and providing a stable initialization for reinforcement learning (RL). We further propose Constrained and Hierarchical Ratio Policy Optimization (CHRPO) to explicitly incentivize question-solving ability under lower budgets by a hierarchical reward. Experiments on three mathematical reasoning benchmarks show the superiority of Extra-CoT. For example, on MATH-500 using Qwen3-1.7B, Extra-CoT achieves over 73\% token reduction with an accuracy improvement of 0.6\%, significantly outperforming state-of-the-art (SOTA) methods. Our source codes have been released at https://github.com/Mwie1024/Extra-CoT.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.15932

Beyond NL2Code: A Structured Survey of Multimodal Code Intelligence

作者:

Xuanle Zhao↗Qiushi Sun↗Jingyu Xiao↗Xuexin Liu↗Haoyue Yang↗Qiaosheng Chen↗Xianzhen Luo↗Jing Huang↗Yufeng Zhong↗Lei Chen↗Shuai Fu↗Zhenlin Wei↗…

While LLMs have substantially advanced text-to-code synthesis, many real programming tasks specify intent through visual artifacts such as screenshots, charts, documents, vector drawings, videos, and interactive states. These tasks require models to connect visual perception to executable programs, because correctness depends not only on syntax but also on layout, geometry, data semantics, editability, interaction behavior, and domain-specific constraints that apply after execution. This survey examines Multimodal Code Intelligence, covering systems that generate, edit, refine, execute, or reason with code under visually grounded inputs and outputs. We first formulate the field by the role that code plays in each task, distinguishing code as a rendered artifact, an editable symbolic structure, a scientific representation, an intermediate reasoning trace, or an executable policy or tool interface. We then organize benchmarks and methods into four domains: Graphical User Interface, Scientific Visualization, Structured Graphics, and Frontier Tasks and Frameworks. This taxonomy connects mature artifact-generation problems to emerging agentic and unified settings and allows us to compare how different tasks treat evidence of correctness. Looking ahead, we argue that future research may benefit from four verification-centered directions. Multi-signal validation can combine complementary evidence of correctness, multi-state verification can test behavior across execution trajectories, cross-task transfer testing can probe reusable visual-code skills, and verifiable agent traces can reveal whether agent actions are grounded in visual evidence. Together, these directions may move multimodal code generation from single-output imitation toward evidence-grounded executable systems.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.12634

Keep Policy Gradient in Charge: Sibling-Guided Credit Distillation for Long-Horizon Tool-Use Agents

作者:

Tianyu Ding↗Jianhong Xin↗Juan Pablo De la Cruz Weinstein↗

Long-horizon tool-use reinforcement learning can learn from outcome verification, but its trajectory-level advantage is broadcast across many reasoning, API, and answer tokens. Self-distillation promises a denser signal by reusing a policy's own rollouts or a privileged teacher. We show, however, that direct token-level self-distillation can silently destroy tool use: it rehearses teacher behavior without knowing which actions the verifier rewards, so useful skills and harmful shortcuts are amplified together. We introduce Sibling-Guided Credit Distillation (SGCD), which uses distillation for credit assignment rather than as a competing actor loss. Dynamic sampling produces mixed successful and failed sibling rollouts; an external LLM summarizes their contrast into a training-only stepwise credit reference; dense teacher/student divergence drives credit reassignment; and bounded detached credit weights reshape GRPO token advantages. The deployed student sees no external LLM, sibling evidence, or oracle. Across AppWorld and $\tau^3$-airline, SGCD improves over matched GRPO comparators: AppWorld TGC $42.9 \to 45.6$ on test_normal and $24.7 \to 27.0$ on test_challenge, and $\tau^3$-airline pass@1 $0.583 \to 0.602$.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.19256

X+Slides: Benchmarking Audience-Conditioned Slide Generation

作者:

Haodong Chen↗Xuanhe Zhou↗Wei Zhou↗Xinyue Shao↗Yanbing Zhu↗Bo Wang↗Jiawei Hong↗Anya Jia↗Fan Wu↗

arXiv:2606.19256v1 Announce Type: new Abstract: Automatically generating slide decks from source documents is an important application of large language models (LLMs). Existing benchmarks primarily assess slide completeness and technical depth, while overlooking the target audience as a critical real-world factor. For instance, specialists demand rigorous proofs, whereas decision-makers prioritize actionable conclusions. To bridge this gap, we introduce X+Slides, a benchmark specifically designed for audience-conditioned slide generation. Built on a diverse corpus spanning 113 topics and seven presentation scenes, X+Slides employs a dynamic evaluation framework constructed from 8,133 deduplicated, source-grounded probes. By assigning audience-specific utility weights to the same source-grounded probes, X+Slides reports four complementary metrics: Audience Coverage measures how much audience-essential information is conveyed, Domain-wise Coverage shows which information types are covered, Efficiency measures delivered utility per unit of attention cost, and Correctness verifies whether slide claims are supported by the source. Experiments on DeepPresenter, SlideTailor, and NotebookLM show that current systems can recover a substantial but still incomplete part of audience-essential information: at $\tau_A=0.7$, DeepPresenter reaches a best Audience Coverage of 0.714, SlideTailor reaches 0.594, and the NotebookLM ablation reaches 0.853 while showing clear grounding differences. These results indicate that visual quality and broad topic coverage should not be treated as evidence support without source-grounded evaluation.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.CV) 2026-06-18 DOI: arXiv:2606.19338

Beyond the Current Observation: Evaluating Multimodal Large Language Models in Controllable Non-Markov Games

作者:

Shengyuan Ding↗Xilin Wei↗Xinyu Fang↗Haodong Duan↗Dahua Lin↗Jiaqi Wang↗Yuhang Zang↗

Deploying multimodal foundation models as closed-loop policies increasingly requires conditioning actions on observations that are no longer visible. However, existing benchmarks either expose the full state, conflate hidden-state reconstruction with other agent skills, or test recall only after an episode has ended. We introduce RNG-Bench (Reconstructive Non-Markov Games), a benchmark suite designed to isolate a base model's ability to reconstruct past observations and act on them during multi-step interaction. RNG-Bench includes two complementary games: Matching Pairs, where card identities briefly revealed at specific locations must later be recalled, and 3D Maze, where egocentric views must be integrated into a spatial map. Both games are evaluated under a unified harness with three controlled difficulty axes: grid size, visual pattern, and observation modality. The benchmark further introduces a head-to-head duel protocol to control for instance-level variance and a Memory Gap metric that disentangles forgetting from poor action selection. The hardest configurations require contexts of roughly 128K tokens and 350 image inputs per episode, and remain far from saturated by frontier MLLMs. Memory Gap analysis shows that most residual errors stem from forgetting earlier observations rather than from suboptimal decision making. Finally, fine-tuning Qwen3.5-9B on optimal-policy rollouts and filtered model demonstrations improves performance on RNG-Bench and transfers to existing benchmarks without degrading general multimodal capability.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2606.11573

Understanding Cross-Sensor Feature Variations for Generalizable 3D Perception

作者:

Xin Qiu↗Wenjie Liu↗Fuyuan Ai↗YuChen Tan↗Zhiwei Xu↗Chunyi Song↗

Radar-camera BEV perception often suffers from degraded performance when evaluated across datasets, as changes in driving scenes, sensor configurations, and environmental conditions can alter both the input observations and the internal fused representations. This work studies this issue from the perspective of source-domain variation modeling, aiming to improve the robustness of BEV-based 3D detectors without relying on target-domain samples. We introduce a framework that characterizes visual scene variations in the frequency domain and uses them to synthesize diverse source-domain views. By comparing the resulting fused BEV representations, the framework further captures how image-level variations influence multi-modal BEV features. These variation patterns are then used to regularize the detector, encouraging the learned fusion space to remain stable under latent scene changes. The proposed method is applied only during training and leaves the inference pipeline unchanged. Experiments on cross-dataset radar-camera 3D detection between View-of-Delft and TJ4DRadSet demonstrate consistent improvements over multiple BEV fusion backbones, and the gains remain effective when a small amount of target-domain data is available.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.17109

Timestamp-Aware Spatio-Temporal Graph Contrastive Learning for Network Intrusion Detection

作者:

Jianli Dai↗Guangwei Wu↗Jiacheng Li↗Weiping Wang↗An He↗Xinjun Xiao↗

arXiv:2606.17109v1 Announce Type: cross Abstract: Given their effectiveness in modeling the relational structure among network traffic flows, graph neural networks (GNNs) have been widely adopted in network intrusion detection systems (NIDSs). However, most existing GNN-based NIDS approaches focus on the relational structure of traffic flows, and treat them as temporally independent, which limits their ability to cope with evolving attack behaviors. Moreover, their reliance on supervised or semi-supervised learning often restricts generalization to unseen attacks. To address these limitations, we propose a novel self-supervised GNN-based framework. To the best of our knowledge, the proposed model is among the first self-supervised GNN-based NIDS models to explicitly leverage real timestamps, which provides faithful temporal dependencies for representation learning. We first construct a series of temporal graphs from network traffic flows according to their timestamps, and then employ an E-GraphSAGE and LSTM based encoder to fully extract temporal information and spatial dependencies of network traffic, without introducing time-costly attention mechanisms. A multi-view graph contrastive learning (GCL) scheme is introduced, where temporal, spatial, and feature contrasts are jointly performed to capture temporal continuity, preserve structural consistency, and improve the generalization and robustness of the learned representations, respectively. In addition, a gradient-norm-based adaptive weighting strategy is designed to optimize the contrastive loss weights. Experimental results on four representative NIDS datasets with real timestamps demonstrate that our method significantly outperforms existing self-supervised approaches and achieves performance comparable to the supervised state-of-the-art GNN method, while maintaining high computational efficiency.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.CV) 2026-06-15 DOI: arXiv:2606.13919

GMN4AD: Graph Matching Network for Alzheimer's Disease Diagnosis with Test-Time Domain Adaptation using Multi-centered Structure Magnetic Resonance Imaging

作者:

Chen Zhao↗Huan Huang↗Yixin Xie↗Jiajing Huang↗Weihua Zhou↗Nandakumar Narayanan↗

Alzheimer's Disease (AD) is a progressive neurodegenerative disorder that affects millions of older adults, with prevalence expected to rise significantly in the coming years. Early diagnosis, particularly during the mild cognitive impairment (MCI) stage, is critical for timely intervention. Structural Magnetic Resonance Imaging (sMRI) has emerged as a key modality for detecting AD-related brain changes, but traditional graph-based approaches often struggle with modality and inter-site heterogeneity, limiting diagnostic performance. In this paper, we propose Graph Matching Network for Alzheimer's Disease Diagnosis (GMN4AD), designed to model interactions between heterogeneous brain graphs derived from neuroimaging data. Unlike conventional methods that treat each brain graph independently, GMN4AD leverages graph matching to capture cross-graph relationships, enhancing diagnostic precision. Furthermore, we introduce a test-time domain adaptation strategy that combines contrastive learning to mitigate domain shifts during inference. Extensive experiments on three public AD datasets demonstrate that GMN4AD achieves superior performance compared to state-of-the-art methods, offering a robust and generalizable solution for AD diagnosis.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18259

Caring Without Feeling: Affective Dynamics as the Control Layer of Human-AI Agent Collaboration

作者:

Junjie Xu↗Xingjiao Wu↗Zihao Zhang↗Yujia Xu↗Yuzhe Yang↗Jin Zhu↗Luwei Xiao↗Wen Wu↗Liang He↗

arXiv:2606.18259v1 Announce Type: cross Abstract: AI agents that plan, retain memory across sessions, invoke external tools and act with partial autonomy are transforming human–AI collaboration. Research on affective computing, simulated empathy in large language models, trust in automation and AI safety has illuminated important design principles, yet these literatures remain fragmented. No integrated account explains how affective cues operate within agentic collaboration – settings in which humans delegate, monitor and correct consequential tasks. This Review synthesises computational and interactional mechanisms of affective dynamics: the processes through which affective cues, emotion-like behaviour and perceived agent affect shape trust calibration, delegation decisions, error correction, dependence and governance. We trace how model-generated affective signals enter interaction loops that govern reliance, repair and oversight, and propose a framework that treats affect not as an internal property of AI but as a coordination layer through which humans and agents negotiate capability, uncertainty and responsibility. The framework provides a foundation for calibrated measurement, purposeful design and informed governance.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16175

PAL-Bench: Evidence-Grounded Profile Reconstruction from Longitudinal Personal Albums

作者:

Qiwei Yan↗Zhiqiang Yuan↗Zexi Jia↗Nanxing Hu↗Kailin Lyu↗Jie Zhou↗Jinchao Zhang↗

arXiv:2606.16175v1 Announce Type: new Abstract: Longitudinal personal albums are weak-schema multimodal databases: noisy perceptual records whose key facts require joins across faces, text, timestamps, locations, and repeated events. Existing visual, video, document, and lifelog benchmarks test sub-problems, but not album-scale profile reconstruction with social identity binding and evidence citation. Benchmarking this task is difficult because the ground truth needed for evaluation–owner profiles, social graphs, face-name maps, and evidence provenance–is private state that real albums cannot safely release. We introduce PAL-Bench, a controlled benchmark for evidence-grounded reconstruction under a public-record contract. Its Evidence Compiler builds latent private worlds, programs target-level evidence paths, renders album pixels, re-measures them through perception pipelines, and exports audited public/private views. Agents receive only perception-derived public records; targets, identifier maps, and evidence paths remain hidden. PAL-Bench contains 50 synthetic users, 36,659 public photo records, and 2,799 targets over owner facts, identities, and relations. A privacy-preserving audit with 10 participants confirms that PAL-Bench evidence structures match real private albums, though equivalent releases remain privacy-prohibitive. Across seven systems and two compute-matched diagnostics, a seven-metric protocol reveals a gap between plausible profile summarization and faithful social reconstruction: systems recover some owner facts but struggle with recurring identities and evidence citation. PAL-TRACE, a reference framework that freezes identity bindings before owner-fact mining, performs best but leaves hard identity resolution far from solved. PAL-Bench provides a testbed for perceptual entity resolution, multimodal data integration, temporal evidence aggregation, and provenance-aware structured prediction.

阅读与讨论 → 访问原文 →