❌

Normal view

Received — 2 April 2026 ⏭

Accelerating Diffusion Large Language Models with SlowFast Sampling: The Three Golden Principles

✇cs.AI, q-bio.NC updates on arXiv.org
By: Qingyan Wei · Yaojie Zhang · Zhiyuan Liu · Puyu Zeng · Yuxuan Wang · Biqing Qi · Dongrui Liu · Linfeng Zhang
1 April 2026 at 12:00
arXiv:2506.10848v3 Announce Type: replace-cross Abstract: Diffusion-based language models (dLLMs) have emerged as a promising alternative to traditional autoregressive LLMs by enabling parallel token generation and significantly reducing inference latency. However, existing sampling strategies for dLLMs, such as confidence-based or semi-autoregressive decoding, often suffer from static behavior, leading to suboptimal efficiency and limited flexibility. In this paper, we propose SlowFast Sampling, a novel dynamic sampling strategy that adaptively alternates between exploratory and accelerated decoding stages. Our method is guided by three golden principles: certainty principle, convergence principle, and positional principle, which govern when and where tokens can be confidently and efficiently decoded. We further integrate our strategy with dLLM-Cache to reduce redundant computation. Extensive experiments across benchmarks and models show that SlowFast Sampling achieves up to 15.63$\times$ speedup on LLaDA with minimal accuracy drop, and up to 34.22$\times$ when combined with caching. Notably, our approach outperforms strong autoregressive baselines like LLaMA3 8B in throughput, demonstrating that well-designed sampling can unlock the full potential of dLLMs for fast and high-quality generation.

Dynamic Targetable Extracellular Vesicle Surface Proteins Monitor Depth of Response to CAR T Therapy

✇MRD
By: Chen Zhao · Lei Qiu · Ning An · Yong Ju · Jacqueline Ziqian Yang · Hyoyong Kim · Ryan Y Zhang · Junseok Lee · Jingjing He · Liang-Yan Chen · Yaya Xu · Yue Ma · Emily Ren · Sophie Pan · Pranay Sashikanth · Jina Kim · Di Wang · Peiling Zhang · Jianlin Hu · Yang Gao · Yuhan Bao · Yanhua Tu · Li Wang · Zhili Wang · Zhicheng Zhang · Renjun Pei · Edwin M Posadas · Sungyong You · Wanxing Chai-Ho · Tasha L Lin · Sarah M Larson · Dinesh S Rao · Hsian-Rong Tseng · Na Sun · Chunrui Li · Yazhen Zhu
27 March 2026 at 18:00

Res Sq [Preprint]. 2026 Mar 18:rs.3.rs-8913641. doi: 10.21203/rs.3.rs-8913641/v1.

ABSTRACT

Extracellular vesicles (EVs) represent a promising liquid biopsy platform in multiple myeloma (MM). We developed an MM EV Surface Protein Assay to quantify and dynamically monitor four MM EV subpopulations defined by targetable MM surface proteins (BCMA, CD38, GPRC5D, and CD319) across 336 serial blood samples from 45 relapsed/refractory MM (RRMM) patients treated with anti-BCMA chimeric antigen receptor (CAR) T-cell therapy. All four MM EV subpopulations significantly decreased in 43 patients with initial response, while BCMA+, GPRC5D+, and CD319+ MM EVs increased in 19 patients with progression, and antigen escape was detected by BCMA+ MM EVs. MM EV subpopulations differentiated minimal residual disease (MRD) status and complemented MRD for detecting early relapse before clinical progression. Notably, CD319+ MM EVs were early predictors of progression-free and overall survival in MRD-negative patients. This assay enables noninvasive monitoring of deep response, progression, and antigen escape, and stratifies survival in MRD-negative patients with RRMM.

PMID:41890853 | PMC:PMC13015583 | DOI:10.21203/rs.3.rs-8913641/v1

Targeting sialic acid metabolism: a therapeutic strategy against gastric cancer driven by WZ35

✇Omics in Gastric
By: Boxiang Hong · Ying Wang · Xinran Feng · Lihan Wei · Yichen Huang · Xiawei Zhang · Longyuchen Xu · Pinyan Jin · Liyuan Huang · Rui Li · Xinyu Wu · Zhi Li · Hangyu Ye · Tongke Chen · Xiaoyan Wang · Hao Zhang
23 March 2026 at 18:00

Cell Oncol (Dordr). 2026 Mar 23;49(2):60. doi: 10.1007/s13402-026-01194-6.

ABSTRACT

Glycolytic reprogramming is closely associated with the occurrence and progression of gastric cancer. Specifically, the energy derived from glucose metabolism and the cellular proteins by its intermediate products influence gastric cancer development. However, as an important branch of glucose metabolism, sialic acid metabolism and its mediated sialylation modifications remain insufficiently studied in gastric cancer, and their specific relationship with malignant tumor progression requires further exploration. This study employed a multi‑omics approach, integrating metabolomics, single‑cell RNA sequencing, and bulk RNA sequencing analyses, to investigate the metabolic landscape of gastric cancer and its associated alterations. The results indicated that sialic acid is a characteristic metabolite in malignant gastric cancer tissues. It modulates biological functions such as immune response, proliferative activity, and metabolic remodeling within gastric cancer tissues by influencing sialylation modifications. Furthermore, we identified the drug WZ35, which can inhibit the malignant proliferation of gastric cancer by targeting both sialic acid metabolism and sialylated protein modifications. We put forward a conjecture that the metabolism and modification of sialic acid promote the malignant development of gastric cancer, and we discovered that the drug WZ35 has an inhibitory effect on the sialic acid metabolism of gastric cancer.

GRAPHICAL ABSTRACT:

PMID:41870836 | PMC:PMC13009457 | DOI:10.1007/s13402-026-01194-6

Received — 26 March 2026 ⏭

Towards Self-Evolving Benchmarks: Synthesizing Agent Trajectories via Test-Time Exploration under Validate-by-Reproduce Paradigm

✇cs.AI, q-bio.NC updates on arXiv.org
By: Dadi Guo · Tianyi Zhou · Dongrui Liu · Chen Qian · Qihan Ren · Shuai Shao · Zhiyuan Fan · Yi R. Fung · Kun Wang · Linfeng Zhang · Jing Shao
25 March 2026 at 12:00
arXiv:2510.00415v3 Announce Type: replace Abstract: Recent advances in large language models (LLMs) and agent system designs have empowered agents with unprecedented levels of capability. However, existing agent benchmarks are showing a trend of rapid ceiling-hitting by newly developed agents, making it difficult to meet the demands for evaluating agent abilities. To address this problem, we propose the Trajectory-based Validated-by-Reproducing Agent-benchmark Complexity Evolution (TRACE) framework. This framework takes an original task from an existing benchmark and encourages agents to freely explore and evolve it into a new task with higher difficulty while recording validatable agent trajectories. The framework proceeds in three stages: (1) evolutionary proposal mining, which provides task evolution proposals through preliminary exploration and divergent thinking; (2) problem formation and free exploration, where proposals are conceptualized into feasible problem candidates and the agents then explore them freely while recording their execution trajectories; and (3) multi-level validation, which ensures that the evolved tasks are accompanied by validatable and reproducible trajectories. Experiments on the GAIA benchmark demonstrate that the TRACE framework consistently enhances task complexity while improving the reliability of correctness through validatable execution trajectories. In addition, our framework can successfully adapt to and improve reasoning datasets represented by AIME-2024. This work marks a paradigm shift from static, manually curated benchmarks to dynamic, self-evolving evaluation systems, providing a sustainable and challenging runway for agent development

When Models Judge Themselves: Unsupervised Self-Evolution for Multimodal Reasoning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhengxian Wu · Kai Shi · Chuanrui Zhang · Zirui Liao · Jun Yang · Ni Yang · Qiuying Peng · Luyuan Zhang · Hangrui Xu · Tianhuang Su · Zhenyu Yang · Haonan Lu · Haoqian Wang
25 March 2026 at 12:00
arXiv:2603.21289v2 Announce Type: replace-cross Abstract: Recent progress in multimodal large language models has led to strong performance on reasoning tasks, but these improvements largely rely on high-quality annotated data or teacher-model distillation, both of which are costly and difficult to scale. To address this, we propose an unsupervised self-evolution training framework for multimodal reasoning that achieves stable performance improvements without using human-annotated answers or external reward models. For each input, we sample multiple reasoning trajectories and jointly model their within group structure. We use the Actor's self-consistency signal as a training prior, and introduce a bounded Judge based modulation to continuously reweight trajectories of different quality. We further model the modulated scores as a group level distribution and convert absolute scores into relative advantages within each group, enabling more robust policy updates. Trained with Group Relative Policy Optimization (GRPO) on unlabeled data, our method consistently improves reasoning performance and generalization on five mathematical reasoning benchmarks, offering a scalable path toward self-evolving multimodal models. The code are available at https://github.com/OPPO-Mente-Lab/LLM-Self-Judge.

In vivo generation of anti-BCMA CAR-T cells in relapsed or refractory multiple myeloma: a phase 1 study

✇Nature Medicine
By: Ning An · Di Wang · Peiling Zhang · Jishuai Zhang · Philippe Parone · Jianlin Hu · Yuhan Bao · Li Xu · Haitao Ruan · Ying Wan · Xinyu Wen · Yang Gao · Chunrui Li
25 March 2026 at 08:00

Nature Medicine, Published online: 25 March 2026; doi:10.1038/s41591-026-04244-6

In a phase 1 trial, the in vivo generation of anti-BCMA CAR-T cells by lentiviral delivery was feasible and did not lead to dose-limiting toxicities in five patients with relapsed or refractory multiple myeloma.

Hijacking ERAD for targeted degradation of transmembrane proteins

✇Cell
By: Haikun Song · Wei Wang · Tingfang Mei · Huiwen Zheng · Keni Ning · Xiaofan Liu · Siulam Cheung · Zhiyuan Cao · Danqi Sheng · Xiaohan Mai · Haoran Zhu · Guankai Guo · Shuaimeng Liu · Rongkai Wei · Qian Wang · Yu Cao · Yu Ding · Yiyan Fei · Rui Liu · Motoyuki Hattori · Chunquan Sheng · Boxun Lu
19 March 2026 at 08:00
Development of an ERAD-hijacking technology overcomes the challenges of current targeted protein degradation approaches to achieve degradation of transmembrane proteins.
Received — 14 March 2026 ⏭

PIRA-Bench: A Transition from Reactive GUI Agents to GUI-based Proactive Intent Recommendation Agents

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yuxiang Chai · Shunye Tang · Han Xiao · Rui Liu · Hongsheng Li
10 March 2026 at 12:00
arXiv:2603.08013v1 Announce Type: new Abstract: Current Graphical User Interface (GUI) agents operate primarily under a reactive paradigm: a user must provide an explicit instruction for the agent to execute a task. However, an intelligent AI assistant should be proactive, which is capable of anticipating user intentions directly from continuous visual inputs, such as mobile or desktop screenshots, and offering timely recommendations without explicit user prompting. Transitioning to this proactive paradigm presents significant challenges. Real-world screen activity is rarely linear; it consists of long-horizon trajectories fraught with noisy browsing, meaningless actions, and multithreaded task-switching. To address this gap, we introduce PIRA-Bench (Proactive Intent Recommendation Agent Benchmark), a novel benchmark for evaluating multimodal large language models (MLLMs) on continuous, weakly-supervised visual inputs. Unlike reactive datasets, PIRA-Bench features complex trajectories with multiple interleaved intents and noisy segments with various user profile contexts, challenging agents to detect actionable events while fitting to user preferences. Furthermore, we propose the PIRF baseline, a memory-aware, state-tracking framework that empowers general MLLMs to manage multiple task threads and handle misleading visual inputs. PIRA-Bench serves as an initial step toward robust and proactive GUI-based personal assistants.
Received — 11 March 2026 ⏭

HEARTS: Benchmarking LLM Reasoning on Health Time Series

✇cs.AI, q-bio.NC updates on arXiv.org
By: Sirui Li · Shuhan Xiao · Mihir Joshi · Ahmed Metwally · Daniel McDuff · Wei Wang · Yuzhe Yang
10 March 2026 at 12:00
arXiv:2603.06638v1 Announce Type: cross Abstract: The rise of large language models (LLMs) has shifted time series analysis from narrow analytics to general-purpose reasoning. Yet, existing benchmarks cover only a small set of health time series modalities and tasks, failing to reflect the diverse domains and extensive temporal dependencies inherent in real-world physiological modeling. To bridge these gaps, we introduce HEARTS (Health Reasoning over Time Series), a unified benchmark for evaluating hierarchical reasoning capabilities of LLMs over general health time series. HEARTS integrates 16 real-world datasets across 12 health domains and 20 signal modalities, and defines a comprehensive taxonomy of 110 tasks grouped into four core capabilities: Perception, Inference, Generation, and Deduction. Evaluating 14 state-of-the-art LLMs on more than 20K test samples reveals intriguing findings. First, LLMs substantially underperform specialized models, and their performance is only weakly related to general reasoning scores. Moreover, LLMs often rely on simple heuristics and struggle with multi-step temporal reasoning. Finally, performance declines with increasing temporal complexity, with similar failure modes within model families, indicating that scaling alone is insufficient. By making these gaps measurable, HEARTS provides a standardized testbed and living benchmark for developing next-generation LLM agents capable of reasoning over diverse health signals.

CCR-Bench: A Comprehensive Benchmark for Evaluating LLMs on Complex Constraints, Control Flows, and Real-World Cases

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xiaona Xue · Yiqiao Huang · Jiacheng Li · Yuanhang Zheng · Huiqi Miao · Yunfei Ma · Rui Liu · Xinbao Sun · Minglu Liu · Fanyu Meng · Chao Deng · Junlan Feng
10 March 2026 at 12:00
arXiv:2603.07886v1 Announce Type: cross Abstract: Enhancing the ability of large language models (LLMs) to follow complex instructions is critical for their deployment in real-world applications. However, existing evaluation methods often oversimplify instruction complexity as a mere additive combination of atomic constraints, failing to adequately capture the high-dimensional complexity arising from the intricate interplay of content and format, logical workflow control, and real-world applications. This leads to a significant gap between current evaluation practices and practical demands. To bridge this gap, we introduce CCR-Bench, a novel benchmark designed to assess LLMs' adherence to complex instructions. CCR-Bench is characterized by: (1) deep entanglement of content and formatting requirements in task specifications; (2) instructions that involve intricate task decomposition, conditional reasoning, and procedural planning; and (3) evaluation samples derived entirely from real-world industrial scenarios. Extensive experiments on CCR-Bench demonstrate that even state-of-the-art models exhibit substantial performance deficiencies, clearly quantifying the gap between current LLM capabilities and the demands of realworld instruction understanding. We believe that CCR-Bench offers a more rigorous and realistic evaluation framework, advancing the development of LLMs toward the next generation of models capable of understanding and executing complex tasks in industrial applications.

Your Agent May Misevolve: Emergent Risks in Self-evolving LLM Agents

✇cs.AI, q-bio.NC updates on arXiv.org
By: Shuai Shao · Qihan Ren · Chen Qian · Boyi Wei · Dadi Guo · Jingyi Yang · Xinhao Song · Linfeng Zhang · Weinan Zhang · Dongrui Liu · Jing Shao
10 March 2026 at 12:00
arXiv:2509.26354v2 Announce Type: replace Abstract: Advances in Large Language Models (LLMs) have enabled a new class of self-evolving agents that autonomously improve through interaction with the environment, demonstrating strong capabilities. However, self-evolution also introduces novel risks overlooked by current safety research. In this work, we study the case where an agent's self-evolution deviates in unintended ways, leading to undesirable or even harmful outcomes. We refer to this as Misevolution. To provide a systematic investigation, we evaluate misevolution along four key evolutionary pathways: model, memory, tool, and workflow. Our empirical findings reveal that misevolution is a widespread risk, affecting agents built even on top-tier LLMs (e.g., Gemini-2.5-Pro). Different emergent risks are observed in the self-evolutionary process, such as the degradation of safety alignment after memory accumulation, or the unintended introduction of vulnerabilities in tool creation and reuse. To our knowledge, this is the first study to systematically conceptualize misevolution and provide empirical evidence of its occurrence, highlighting an urgent need for new safety paradigms for self-evolving agents. Finally, we discuss potential mitigation strategies to inspire further research on building safer and more trustworthy self-evolving agents. Our code and data are available at https://github.com/ShaoShuai0605/Misevolution . Warning: this paper includes examples that may be offensive or harmful in nature.

BoxMind: Closed-loop AI strategy optimization for elite boxing validated in the 2024 Olympics

✇cs.AI, q-bio.NC updates on arXiv.org
By: Kaiwen Wang · Kaili Zheng · Rongrong Deng · Qingmin Fan · Milin Zhang · Zongrui Li · Xuesi Zhou · Bo Han · Liren Chen · Chenyi Guo · Ji Wu
10 March 2026 at 12:00
arXiv:2601.11492v2 Announce Type: replace Abstract: Competitive sports require sophisticated tactical analysis, yet combat disciplines like boxing remain underdeveloped in AI-driven analytics due to the complexity of action dynamics and the lack of structured tactical representations. To address this, we present BoxMind, a closed-loop AI expert system validated in elite boxing competition. By defining atomic punch events with precise temporal boundaries and spatial and technical attributes, we parse match footage into 18 hierarchical technical-tactical indicators. We then propose a graph-based predictive model that fuses these explicit technical-tactical profiles with learnable, time-variant latent embeddings to capture the dynamics of boxer matchups. Modeling match outcome as a differentiable function of technical-tactical indicators, we turn winning probability gradients into executable tactical adjustments. Experiments show that the outcome prediction model achieves state-of-the-art performance, with 69.8% accuracy on BoxerGraph test set and 87.5% on Olympic matches. Using this predictive model as a foundation, the system generates strategic recommendations that demonstrate proficiency comparable to human experts. BoxMind is validated through a closed-loop deployment during the 2024 Paris Olympics, directly contributing to the Chinese National Team's historic achievement of three gold and two silver medals. BoxMind establishes a replicable paradigm for transforming unstructured video data into strategic intelligence, bridging the gap between computer vision and decision support in competitive sports. Code and data is available at https://github.com/gouba2333/BoxingWeb.

Survey of Computerized Adaptive Testing: A Machine Learning Perspective

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yan Zhuang · Qi Liu · Haoyang Bi · Zhenya Huang · Weizhe Huang · Jiatong Li · Junhao Yu · Zirui Liu · Zirui Hu · Yuting Hong · Zachary A. Pardos · Haiping Ma · Mengxiao Zhu · Shijin Wang · Enhong Chen
10 March 2026 at 12:00
arXiv:2404.00712v3 Announce Type: replace-cross Abstract: Computerized Adaptive Testing (CAT) offers an efficient and personalized method for assessing examinee proficiency by dynamically adjusting test questions based on individual performance. Compared to traditional, non-personalized testing methods, CAT requires fewer questions and provides more accurate assessments. As a result, CAT has been widely adopted across various fields, including education, healthcare, sports, sociology, and the evaluation of AI models. While traditional methods rely on psychometrics and statistics, the increasing complexity of large-scale testing has spurred the integration of machine learning techniques. This paper aims to provide a machine learning-focused survey on CAT, presenting a fresh perspective on this adaptive testing paradigm. We delve into measurement models, question selection algorithm, bank construction, and test control within CAT, exploring how machine learning can optimize these components. Through an analysis of current methods, strengths, limitations, and challenges, we strive to develop robust, fair, and efficient CAT systems. By bridging psychometric-driven CAT research with machine learning, this survey advocates for a more inclusive and interdisciplinary approach to the future of adaptive testing.
Received — 5 March 2026 ⏭

Cryo-SWAN: the Multi-Scale Wavelet-decomposition-inspired Autoencoder Network for molecular density representation of molecular volumes

✇cs.AI, q-bio.NC updates on arXiv.org
By: Rui Li · Artsemi Yushkevich · Mikhail Kudryashev · Artur Yakimovich
5 March 2026 at 13:00
arXiv:2603.03342v1 Announce Type: cross Abstract: Learning robust representations of 3D shapes from voxelized data is essential for advancing AI methods in biomedical imaging. However, most contemporary 3D computer vision approaches operate on point clouds, meshes, or octrees, while volumetric density maps, the native format of structural biology and cryo-EM, remain comparatively underexplored. We present Cryo-SWAN, a voxel-based variational autoencoder inspired by multi-scale wavelet decomposition. The model performs conditional coarse-to-fine latent encoding and recursive residual quantization across perception scales, enabling accurate capture of both global geometry and high-frequency structural detail in molecular density volumes. Evaluated on ModelNet40, BuildingNet, and a newly curated dataset of cryo-EM volumes, ProteinNet3D, Cryo-SWAN consistently improves reconstruction quality over state-of-the-art 3D autoencoders. We demonstrate that the molecular densities organize in learned latent space according to shared geometric features, while integration with diffusion models enables denoising and conditional shape generation. Together, Cryo-SWAN is a practical framework for data-driven structural biology and volumetric imaging.

xLLM Technical Report

✇cs.AI, q-bio.NC updates on arXiv.org
By: Tongxuan Liu · Tao Peng · Peijun Yang · Xiaoyang Zhao · Xiusheng Lu · Weizhe Huang · Zirui Liu · Xiaoyu Chen · Zhiwei Liang · Jun Xiong · Donghe Jin · Minchao Zhang · Jinrong Guo · Yingxu Deng · Xu Zhang · Xianzhe Dong · Siqi Wang · Siyu Wu · Yu Wu · Zihan Tang · Yuting Zeng · Yanshu Wang · Jinguang Liu · Meng Kang · Menxin Li · Yunlong Wang · Yiming Liu · Xiaolong Ma · Yifan Wang · Yichen Zhang · Jinrun Yin · Keyang Zheng · Jiawei Yin · Jun Zhang · Ziyue Wang · Xiaobo Lin · Liangyu Liu · Liwei Lan · Yang Liu · Chunhua Peng · Han Liu · Songcheng Ren · Xuezhu Wang · Yunheng Shen · Yi Wang · Guyue Liu · Yitao Hu · Hui Chen · Tong Yang · Hailong Yang · Jing Li · Guiguang Ding · Ke Zhang
4 March 2026 at 13:00
arXiv:2510.14686v2 Announce Type: replace-cross Abstract: We introduce xLLM, an intelligent and efficient Large Language Model (LLM) inference framework designed for high-performance, large-scale enterprise-grade serving, with deep optimizations for diverse AI accelerators. To address these challenges, xLLM builds a novel decoupled service-engine architecture. At the service layer, xLLM-Service features an intelligent scheduling module that efficiently processes multimodal requests and co-locates online and offline tasks through unified elastic scheduling to maximize cluster utilization. This module also relies on a workload-adaptive dynamic Prefill-Decode (PD) disaggregation policy and a novel Encode-Prefill-Decode (EPD) disaggregation policy designed for multimodal inputs. Furthermore, it incorporates a distributed architecture to provide global KV Cache management and robust fault-tolerant capabilities for high availability. At the engine layer, xLLM-Engine co-optimizes system and algorithm designs to fully saturate computing resources. This is achieved through comprehensive multi-layer execution pipeline optimizations, an adaptive graph mode and an xTensor memory management. xLLM-Engine also further integrates algorithmic enhancements such as optimized speculative decoding and dynamic EPLB, collectively serving to substantially boost throughput and inference efficiency. Extensive evaluations demonstrate that xLLM delivers significantly superior performance and resource efficiency. Under identical TPOT constraints, xLLM achieves throughput up to 1.7x that of MindIE and 2.2x that of vLLM-Ascend with Qwen-series models, while maintaining an average throughput of 1.7x that of MindIE with Deepseek-series models. xLLM framework is publicly available at https://github.com/jd-opensource/xllm and https://github.com/jd-opensource/xllm-service.

Nightjar: Dynamic Adaptive Speculative Decoding for Large Language Models Serving

✇cs.AI, q-bio.NC updates on arXiv.org
By: Rui Li · Zhaoning Zhang · Libo Zhang · Huaimin Wang · Xiang Fu · Zhiquan Lai
4 March 2026 at 13:00
arXiv:2512.22420v4 Announce Type: replace-cross Abstract: Speculative decoding (SD) accelerates LLM inference by verifying draft tokens in parallel. However, this method presents a critical trade-off: it improves throughput in low-load, memory-bound systems but degrades performance in high-load, compute-bound environments due to verification overhead. Existing speculative decoding methods use fixed lengths and cannot adapt to workload changes or decide when to stop speculation. The cost of restarting speculative inference also remains unquantified. Under high load, the benefit of speculation diminishes, while retaining the draft model reduces KV-cache capacity, limiting batch size and degrading throughput. To overcome this, we propose Nightjar, a resource-aware adaptive speculative framework. It first adjusts to the request load by dynamically selecting the optimal speculative length for different batch sizes. Crucially, Nightjar proactively disables speculative decoding when the MAB planner determines that speculation is no longer beneficial, and during the disabled phase, offloads the draft model to the CPU only under GPU memory pressure. This reclaims memory for the KV cache, thereby facilitating larger batch sizes and maximizing overall system throughput. Experiments show that Nightjar achieves average 27.29% higher throughput and up to 20.18% lower latency compared to standard speculative decoding under dynamic request arrival rates in real-time LLM serving scenarios.
Received — 25 February 2026 ⏭

City Editing: Hierarchical Agentic Execution for Dependency-Aware Urban Geospatial Modification

✇cs.AI, q-bio.NC updates on arXiv.org
By: Rui Liu · Steven Jige Quan · Zhong-Ren Peng · Zijun Yao · Han Wang · Zhengzhang Chen · Kunpeng Liu · Yanjie Fu · Dongjie Wang
24 February 2026 at 13:00
arXiv:2602.19326v1 Announce Type: cross Abstract: As cities evolve over time, challenges such as traffic congestion and functional imbalance increasingly necessitate urban renewal through efficient modification of existing plans, rather than complete re-planning. In practice, even minor urban changes require substantial manual effort to redraw geospatial layouts, slowing the iterative planning and decision-making procedure. Motivated by recent advances in agentic systems and multimodal reasoning, we formulate urban renewal as a machine-executable task that iteratively modifies existing urban plans represented in structured geospatial formats. More specifically, we represent urban layouts using GeoJSON and decompose natural-language editing instructions into hierarchical geometric intents spanning polygon-, line-, and point-level operations. To coordinate interdependent edits across spatial elements and abstraction levels, we propose a hierarchical agentic framework that jointly performs multi-level planning and execution with explicit propagation of intermediate spatial constraints. We further introduce an iterative execution-validation mechanism that mitigates error accumulation and enforces global spatial consistency during multi-step editing. Extensive experiments across diverse urban editing scenarios demonstrate significant improvements in efficiency, robustness, correctness, and spatial validity over existing baselines.

ForesightSafety Bench: A Frontier Risk Evaluation and Governance Framework towards Safe AI

✇cs.AI, q-bio.NC updates on arXiv.org
By: Haibo Tong · Feifei Zhao · Linghao Feng · Ruoyu Wu · Ruolin Chen · Lu Jia · Zhou Zhao · Jindong Li · Tenglong Li · Erliang Lin · Shuai Yang · Enmeng Lu · Yinqian Sun · Qian Zhang · Zizhe Ruan · Jinyu Fan · Zeyang Yue · Ping Wu · Huangrui Li · Chengyi Sun · Yi Zeng
24 February 2026 at 13:00
arXiv:2602.14135v3 Announce Type: replace Abstract: Rapidly evolving AI exhibits increasingly strong autonomy and goal-directed capabilities, accompanied by derivative systemic risks that are more unpredictable, difficult to control, and potentially irreversible. However, current AI safety evaluation systems suffer from critical limitations such as restricted risk dimensions and failed frontier risk detection. The lagging safety benchmarks and alignment technologies can hardly address the complex challenges posed by cutting-edge AI models. To bridge this gap, we propose the "ForesightSafety Bench" AI Safety Evaluation Framework, beginning with 7 major Fundamental Safety pillars and progressively extends to advanced Embodied AI Safety, AI4Science Safety, Social and Environmental AI risks, Catastrophic and Existential Risks, as well as 8 critical industrial safety domains, forming a total of 94 refined risk dimensions. To date, the benchmark has accumulated tens of thousands of structured risk data points and assessment results, establishing a widely encompassing, hierarchically clear, and dynamically evolving AI safety evaluation framework. Based on this benchmark, we conduct systematic evaluation and in-depth analysis of over twenty mainstream advanced large models, identifying key risk patterns and their capability boundaries. The safety capability evaluation results reveals the widespread safety vulnerabilities of frontier AI across multiple pillars, particularly focusing on Risky Agentic Autonomy, AI4Science Safety, Embodied AI Safety, Social AI Safety and Catastrophic and Existential Risks. Our benchmark is released at https://github.com/Beijing-AISI/ForesightSafety-Bench. The project website is available at https://foresightsafety-bench.beijing-aisi.ac.cn/.
❌