数据来源：HuggingFace Papers Latest Papers1. AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and SecurityThe rise of AI agents introduces complex safety and security challenges arising from autonomous tool use and environmental interactions. Current guardrail models lack agentic risk awareness and transparency in risk diagnosis. To introduce an agentic guardrail that covers complex and numerous risky behaviors, we first propose a unified three-dimensional taxonomy that orthogonally c ...

数据来源：HuggingFace Papers Latest Papers1. Harder Is Better: Boosting Mathematical Reasoning via Difficulty-Aware GRPO and Multi-Aspect Question ReformulationReinforcement Learning with Verifiable Rewards (RLVR) offers a robust mechanism for enhancing mathematical reasoning in large models. However, we identify a systematic lack of emphasis on more challenging questions in existing methods from both algorithmic and data perspectives, despite their importance for refining underdeveloped capabiliti ...

数据来源：HuggingFace Papers Latest Papers1. Idea2Story: An Automated Pipeline for Transforming Research Concepts into Complete Scientific NarrativesAutonomous scientific discovery with large language model (LLM)-based agents has recently made substantial progress, demonstrating the ability to automate end-to-end research workflows. However, existing systems largely rely on runtime-centric execution paradigms, repeatedly reading, summarizing, and reasoning over large volumes of scientific literatur ...

数据来源：HuggingFace Papers Latest Papers1. LongCat-Flash-Thinking-2601 Technical ReportWe introduce LongCat-Flash-Thinking-2601, a 560-billion-parameter open-source Mixture-of-Experts (MoE) reasoning model with superior agentic reasoning capability. LongCat-Flash-Thinking-2601 achieves state-of-the-art performance among open-source models on a wide range of agentic benchmarks, including agentic search, agentic tool use, and tool-integrated reasoning. Beyond benchmark performance, the model demons ...

数据来源：HuggingFace Papers Latest Papers1. Idea2Story: An Automated Pipeline for Transforming Research Concepts into Complete Scientific NarrativesAutonomous scientific discovery with large language model (LLM)-based agents has recently made substantial progress, demonstrating the ability to automate end-to-end research workflows. However, existing systems largely rely on runtime-centric execution paradigms, repeatedly reading, summarizing, and reasoning over large volumes of scientific literatur ...

数据来源：HuggingFace Papers Latest Papers1. Can LLMs Clean Up Your Mess? A Survey of Application-Ready Data Preparation with LLMsData preparation aims to denoise raw datasets, uncover cross-dataset relationships, and extract valuable insights from them, which is essential for a wide range of data-centric applications. Driven by (i) rising demands for application-ready data (e.g., for analytics, visualization, decision-making), (ii) increasingly powerful LLM techniques, and (iii) the emergence of i ...

数据来源：HuggingFace Papers Latest Papers1. ASTRA: Automated Synthesis of agentic Trajectories and Reinforcement ArenasLarge language models (LLMs) are increasingly used as tool-augmented agents for multi-step decision making, yet training robust tool-using agents remains challenging. Existing methods still require manual intervention, depend on non-verifiable simulated environments, rely exclusively on either supervised fine-tuning (SFT) or reinforcement learning (RL), and struggle with stable lo ...

数据来源：HuggingFace Papers Latest Papers1. Idea2Story: An Automated Pipeline for Transforming Research Concepts into Complete Scientific NarrativesAutonomous scientific discovery with large language model (LLM)-based agents has recently made substantial progress, demonstrating the ability to automate end-to-end research workflows. However, existing systems largely rely on runtime-centric execution paradigms, repeatedly reading, summarizing, and reasoning over large volumes of scientific literatur ...

数据来源：HuggingFace Papers Latest Papers1. CodeOCR: On the Effectiveness of Vision Language Models in Code UnderstandingLarge Language Models (LLMs) have achieved remarkable success in source code understanding, yet as software systems grow in scale, computational efficiency has become a critical bottleneck. Currently, these models rely on a text-based paradigm that treats source code as a linear sequence of tokens, which leads to a linear increase in context length and associated computational c ...

数据来源：微博热搜 排名 话题 热度 分类 1 奋力书写挺膺担当的青春篇章 2 杨幂 妈幂 1480060 3 黄牛为什么能抢走演唱会门票 860851 4 对贵州启动国家三级救灾应急响应 743234 5 杨幂淘宝闪购代言人 6 王嘉尔反驳印度主持人 742444 7 吃荔枝不看的人有福了 434670 8 9平米的房住出了90平方的感觉 358317 9 林志玲一个生煎包吃十几口 354648 10 黄圣依孩子用石明鑫主编的数学题 综艺 353508 11 wakuku 344852 12 今天是平分2025的一天 344473 13 家有儿女小雨变暴雨了 342917 14 优酷爆剧三大赛道全面开花 341700 15 董卿生病但帅 340552 16 你俩长嘴是只会亲嘴吗 剧集 339270 17 银行职员诈骗近亿元打赏主播6千万 338667 18 埃菲尔铁塔被热到弯曲偏斜 338058 19 王楚然求助治寻麻疹的方法 320966 20 ...

数据来源：微博热搜 排名 话题 热度 分类 1 奋力书写挺膺担当的青春篇章 2 一棵荔枝树上竟结出30种荔枝 1164155 3 北京出门像被牛舔了一口 702369 4 网警护航高考志愿填报 634620 5 我放过的坏人杀了我的亲人 剧集 602549 6 女孩高考288分属实 525832 7 知情人称5层楼房垮塌前已倾斜 516046 8 官方辟谣山西榆次遭遇水灾 9 一批抗战主题电视剧将陆续播出 453534 10 吹牛老爹被判组织卖淫罪 438168 11 杭州热到全国第一了 379241 12 中国海军山东舰航母编队抵达香港 373665 13 曝娜扎张云龙近况 369484 14 中文慢慢失去了加密功能 362137 15 2人就餐点茶后还被收11元白开水费 340702 16 亮剑等抗战题材经典作品将展播 339642 17 黄牛为什么能抢走演唱会门票 338710 18 严浩翔安抚黄子韬 336684 19 被麦琳抄袭博主拒绝赔偿只求平 ...

数据来源：微博热搜 排名 话题 热度 分类 1 厦门发展战略启示 2 日本无事发生 1605395 3 鹿晗发现伴舞的屏保是自己 演出 769349 4 高温闷热天气防中暑指南 561557 5 特朗普正式签署大而美法案 457884 6 124斤女子带全家减肥40天共瘦40斤 442413 7 建议不要让任何人的钱过银行卡 429586 8 成都暴雨致家里插座出水成瀑布 419601 9 杨幂20年顶流不是白当的 355055 10 上海一面馆被曝将剩面二次上桌 298572 11 独居老人网购花费200万睡在快递上 249781 12 67岁丈夫出轨50岁闺蜜妻子怒告二人 248350 13 晚晚发视频回应不让助播女生化妆 246262 14 杨幂摔出了神图 244498 15 日本 241625 16 女子3万8卖掉自己的孩子用于打赏主播 239313 17 林志玲说带孩子没有优雅这件事 236821 18 王源求婚转场 232687 19 歌手 郑欣 ...

数据来源：微博热搜 排名 话题 热度 分类 1 总书记和青年朋友在一起 2 马斯克发文成立美国党 1188572 3 杭州东站跳轨事件乘客发声 449371 4 又一次被鱼水情深的双向奔赴感动 352586 5 小伙脖子肿痛开刀取出7颗瓜子 290115 6 马斯克想拿下美国会两院部分席位 280738 7 谁家男主尸体都不放过啊 277109 8 女子指甲下长肿瘤被误诊甲沟炎10年 227899 9 BLG对战T1 220877 10 真有综艺听劝了 204476 11 男子威胁公开前女同事隐私视频 202317 12 男子长期性侵未成年养女被判死刑 202303 13 杭州东站发生卧轨事件 201833 14 rapper用归国四子diss黄子韬 综艺 201550 15 电子垃圾三件套捞中国男人几百亿 201509 16 3名初中生凌晨偷奔驰致严重车祸 201067 17 杭州东站卧轨事件目击者发声 200762 18 建议不要临睡前才刷牙 200647 ...

数据来源：微博热搜 排名 话题 热度 分类 1 中美关系的未来在青年 2 青岛大学凌晨发情况说明 4377481 3 男生与女友同居太兴奋后空翻摔死 2238115 4 文化中国行看古人的清凉好物 1429075 5 舅舅回应16个外甥连续5年来过暑假 931848 6 赛里木湖 857007 7 日军官因找借口挑起卢沟桥事变升职 447346 8 马斯克被警告会成为一个没国家的人 389358 9 七七事变88年了 367993 10 央视曝光零差评背后猫腻 342578 11 张鑫 白鹿站姐 319348 12 青岛大学宿舍图 317755 13 青岛大学学生称有同学宿舍中暑送医 301113 14 重庆一水库水位下降现宋代摩崖造像 282106 15 男子借住朋友家一觉醒来6万没了 278049 16 杨紫脖子上的青筋都哭出来了 剧集 258109 17 鹿晗安慰粉丝别哭别哭 演出 254150 18 山东高温天热得一片红 253810 19 藏海传床 ...

数据来源：HuggingFace Papers Latest Papers1. Green-VLA: Staged Vision-Language-Action Model for Generalist RobotsWe introduce Green-VLA, a staged Vision-Language-Action (VLA) framework for real-world deployment on the Green humanoid robot while maintaining generalization across diverse embodiments. Green-VLA follows a five stage curriculum: (L0) foundational VLMs, (L1) multimodal grounding, (R0) multi-embodiment pretraining, (R1) embodiment-specific adaptation, and (R2) reinforcement-learning (RL) ...