⚠️ 请求失败的平台
- steam
RSS 新增更新
61 条
AI 与前沿
15 条
▸展开看内容
arXiv:2610.00015v1 Announce Type: new Abstract: Large-language-model agents can propose and execute actions, but proposal, authority, dispatch, verified external effect, and serving promotion are different claims. We present Praxa, an agent harness that represents these states explicitly through deterministic admission, brokered execution, external read-back, reconciliation, and reviewed promotion. We report four evidence lanes. First, an author-run repository-local audit at a pinned revision pa
▸展开看内容
arXiv:2610.00074v1 Announce Type: new Abstract: K-Dense BYOK (bring your own keys) is a free, open-source AI research assistant for scientists in any field that runs on the researcher's own computer. The researcher supplies access to a model of their choice, hosted or running locally, and the application supplies everything else: a place for the work to run, a layer of scientific scaffolding, and a complete record. Each project is an ordinary folder, so the data, the code, the results, and the r
▸展开看内容
arXiv:2610.00529v1 Announce Type: new Abstract: The ontology-based contextual AI evaluation (OB-CAIE) methodology was developed to address a lack of scientific rigor that arises from unclear testing coverage, to balance human expertise and automations, and to address a lack of reproducibility of AI evaluation testing environments. OB-CAIE strengthens the current state of AI evaluations by addressing the first step in the scientific method by clearly defining what will be tested. Two ontologies r
▸展开看内容
arXiv:2610.00531v1 Announce Type: new Abstract: AI-generated content, often called AI slop, is increasingly common everywhere, particularly in academia. Slop in AI-generated scientific papers, however, has more complex patterns that cannot be easily detected by existing token-based AI detectors. Each part of such a paper looks plausible while the scientific reasoning that connects the parts breaks down, which can mislead how readers assess the work. We benchmark these failures as scientific slop
▸展开看内容
arXiv:2610.00636v1 Announce Type: new Abstract: Evaluating AI agents on scientific research tasks is constrained by the time and resources required for the underlying experiments or calculations. In computational materials research, repeating the same expensive simulations across agents and trials can make evaluation impractical. We introduce CompMat-Bench, a benchmark of 94 tasks derived from recently published computational materials studies, each asking agents to complete a step toward achiev
▸展开看内容
arXiv:2610.00654v1 Announce Type: new Abstract: Personalization has long relied on customer data to infer what an individual is likely to value. We call this customer evidence: the customer's historical behavior and preferences. Generative AI extends personalization by allowing providers to supply changing situational information at the moment a response is produced, without encoding every condition in advance. We define this provider-side context as information about what is possible, permitted
▸展开看内容
arXiv:2610.00682v1 Announce Type: new Abstract: Large language models (LLMs) increasingly underpin scientific AI applications that reason over structured knowledge, from biomedical question answering to materials informatics. However, their logical reasoning often falls short, producing factual inaccuracies unacceptable in these settings. Reliable evaluation remains challenging: manual dataset construction scales poorly, and LLM-based generation risks embedding the very flaws it aims to measure
▸展开看内容
arXiv:2610.00791v1 Announce Type: new Abstract: Enterprise information is represented through artifacts shaped by applications, projects, technologies, organizational boundaries, and local requirements. These structures accumulate over time, creating representational complexity that must be maintained by the enterprise and interpreted by information consumers and AI systems. This paper introduces Enterprise Representation Simplification (ERS) as reducing unnecessary representational complexity w
▸展开看内容
arXiv:2610.00797v1 Announce Type: new Abstract: Contextual security defenses prevent AI agents from taking rogue actions by synthesizing a task-specific policy and enforcing it on the agent's tool calls. In multi-step tasks, however, which actions are valid often depends on what the agent has already done and learned. We present Sapien, a policy engine for enforcing stateful contextual policies. A Sapien policy specifies permitted tool-call sequences using a regular expression extended with stat
▸展开看内容
华为 Mate90 系列发布,售价 5999 元起 10 月 1 日,华为 Mate90 系列手机正式发布。华为 Mate90 售价 5999 元起,华为 Mate90 Pro 售价 6999 元起,华为 Mate90 Pro Max 售价 9499 元起,华为 Mate90 Pro Max 典藏版售价 10999 元起,华为 Mate90 RS 非凡大师售价 12999 元起。 影像系统是本次迭代的核心亮点。华为 Mate90 系列全系搭载第三代红枫原色摄像头,标配红外光学镀膜技术,有效提升夜景画面纯净度与通透感。新机新增烟花人像、柔光人像、星芒特效等多元拍摄风格,依托原生原片直出的优质画质,既满足日常随手拍需求,也为后期创作预留充足空间。同时,全新华为光变巨炮睿影 Z10 专业影像模块正式登场,区别于传统增距镜设备,其内置 1 英寸大底传感器,支持 1-10 倍连续光学变焦,大幅拓展移动端专业拍摄边界。(来源:36 氪) 苹果首款 OLED 触屏 MacBook Pro 或最快本月发布 苹果首款配备 OLED 显示屏的 MacBook Pro 正在逐渐接近发布。彭博社记者马克·古尔
▸展开看内容
我的“声音”到底归谁?我的“声音”又到底值多少钱?当AI开始轻而易举的模仿我们的声音肆意表达,许多过去从未出现的问题逐一浮现。知名声优、演员、主持人津田健次郎今日起诉tiktok平台引发争议,而在播客圈儿,也不断有我们熟悉的主播声音出现在互联网的视频中,所以我们决定从“声音的价值”这个角度切入,从广播剧的诞生,到声优行业的出现,再到播客时代,被录下的声音如何定价、授权?百年变化背后,藏着和你有关的版权难题,但也似乎是在给ai时代的困境指明了一条出路。【时间轴】02:01 声音背后的争议:从津田健次郎在TIKTOK上的诉讼案例聊起 10:40 探索百年人类声音记录的科技历程… 21:20 日本近现代文学中的“言文一致”运动:平民百姓的文化需求与创作潮流 32:01 声优行业的起点:1925年广播剧研究与日本声优史的关系 42:43 动画片制作背后的艰辛与收益分配:手冢治虫与后续动画行业的博弈 53:24 演员罢工引发的连锁反应:电视台滚动更新受阻,配音行业的重要性逐渐被重视 01:04:03 日本排优联合提出动画重播费用问题,动画配音演员的待遇能否改善? 01:14:44 声音的版权争议
▸展开看内容
一个工程师写完数据处理代码,点下提交,然后去喝了杯咖啡。十几分钟甚至几十分钟之后,他才回到屏幕前看结果、改代码、再提交。这条「等待— 反馈」的空隙,过去一直是数据平台默认的节奏。 Agent 没有这个空隙。它拿到一次结果,立刻验证,紧接着并发地去试第二种、第三种方案,飞快拿到反馈,再发起下一轮调用。 人用平台,发过来的是一个相对确定的任务,算完、给结果、走人;Agent 用平台,会不停地换着法子去完成你交代的那个目标。压力就这样被放大了—— 管控的压力和隐形计算的压力,会十倍、百倍地往上翻。 这正在成为 2026 年云计算需要面对的一种新负载。 AI 正在深入「把模型用起来」的阶段。模型一旦长成 Agent,它就不再是一个躺在那里等你调用的接口,而更像一个会自己找数据、自己拆任务、自己并发试错、自己一直跑下去的数字员工。它制造出来的负载,和过去十几年云平台习惯承载的那种负载,完全是两回事。 阿里云智能集团首席技术官李飞飞在云栖大会上把智能体落地拆成三个词:Model 决定智能的上限,Context 决定 Agent 读不读得懂你的业务,Harness 决定它能不能真正调起工具、把事做完
▸展开看内容
苹果即将发布「智能家居中枢」新品 据彭博社报道,苹果公司计划于 10 月 13 日推进其拖延已久的智能家居市场布局,这是该公司在新任 CEO 约翰·特努斯领导下的一次关键产品扩张。 据知情人士透露,该战略的核心是一款代号为 J490 的智能家居中枢设备。该设备将采用大约 6 英寸的方形显示屏设计,提供可安装在墙上或放置在台面上的版本。 按照苹果的设想,用户会在家中多处放置多台这种显示屏。它们可用于控制恒温器、门锁和其他联网产品,也可用于视频通话、对讲式沟通、播放音乐,以及查看存储在苹果 iCloud 服务中的照片幻灯片。 苹果还把这台设备视为厨房里的通信和信息中枢,用户可以在做饭时通过语音查询菜谱、观看视频或拨打 FaceTime 通话。 同时,报道称苹果也计划宣布 HomePod mini 自 2020 年首次亮相以来的首次更新,以及自 2022 年以来的首款新电视机顶盒。(来源:凤凰网科技) 谷歌 Gemini 4 Argon 登场,长周期代码工程测试超 Opus 5.5 当地时间 9 月 30 日,谷歌发布 Gemini 4 Argon,称其为公司迄今最先进的 AI 模型,但目前
论文与开源
8 条
▸展开看内容
arXiv:2610.00012v1 Announce Type: new Abstract: LLM agents increasingly act through modular systems, such as order, payment, inventory, and shipment services, where actions in one module change which transitions are valid in another. Standard world models usually fit observational traces, but this is not the quantity needed for intervention-time planning: a trace may show that payment precedes shipment without identifying whether payment authorizes shipment, inventory mediates the effect, or a h
▸展开看内容
arXiv:2610.00025v1 Announce Type: new Abstract: Agent harnesses increasingly want to run small language models (SLMs) on the microtasks around a frontier large language model (LLM) planner: auto-approving shell commands, writing memory, selecting tools, ranking past turns. We ask whether off-the-shelf SLMs meet practitioner-defined thresholds and, when they fail, why, and whether quantization changes the answer. We build a benchmark of 4 such microtasks with fixed prompts and automatic metrics
▸展开看内容
arXiv:2610.00313v1 Announce Type: new Abstract: Scientific coding agents receive equations, boundary conditions, and output requirements in writing, then must assess the programs they revise. Rules to Tools (R2T) supplies prepared executable checks of public scientific requirements. Matched SciCode repair groups share written checks, starting programs, model, and budgets; the tool group receives a callable implementation. Across two task-ID cohorts, complete repair is 26/30 with text and 29/30 w
▸展开看内容
arXiv:2610.00349v1 Announce Type: new Abstract: Resource limits are becoming an authorization boundary for AI agents that delegate work across concurrent and failure-prone workers. Parent-child allocation constraints, affine objects, and distributed escrow do not by themselves prevent overspend when replies are lost, effects complete after timeout, messages repeat, branches partition, or DAG joins alias one lineage. We formalize fault-tolerant budget conservation for distributed multi-agent dele
▸展开看内容
arXiv:2610.00366v1 Announce Type: new Abstract: A persistent agent must decide both what to retain as information arrives and what to surface once a query appears, yet memory evaluations can confound these decisions by comparing methods that differ in both retention and selection. We build a streaming-recall benchmark crossing retention and selection rules and evaluate every condition on the same 300 seeded episodes. Holding access fixed, query-aware selection improves required-fact recall by 15
▸展开看内容
arXiv:2610.00372v1 Announce Type: new Abstract: Large language model agents rely on external harnesses to pass information between the model and its environment and to recover from execution errors. Yet recovery is usually judged only by average task success. This hides an important tension. The same operation can rescue a failing trajectory or disrupt one that would otherwise succeed. We frame recovery as a causal decision problem. Starting from the same execution state, we compare what happens
游戏
10 条
▸展开看内容
10月2日,小红书开放世界冒险岛RED LAND正式开幕,上海杨浦复兴岛再次变身”痛岛”。早上不到10点,船台公园入口处已经排起了长队。这是RED LAND连续第二年落地复兴岛,活动将持续至10月6日。相较去年,本届RED LAND规模全面升级:场地从8万平方米扩展至15万平方米,超过160个游戏及动漫IP登岛,双双翻倍。据官方介绍,本届五轮售票平均每轮20.6秒全部售罄,这场持续五天的同好狂欢,也成为今年国庆假期最受关注的青年潮流文化活动之一。 160+个人气IP登岛,打造2.5次元“沉浸式开放世界”走进复兴岛,巨幅人气角色气模簇拥着机械船台,身着各式cos装扮的玩家与岛民NPC穿梭往来,在这里,青年潮流文化与工业遗存交相辉映。本届RED LAND阵容不仅集结了160+人气IP,也集齐了多个“首次”。米哈游旗下四款作品首次线下同框,其生活模拟游戏新作《星布谷地》迎来国内首次线下亮相;复联、X战警、神奇四侠与黑豹五座英雄巨像在《复仇者联盟5》展位全球首亮相,迪士尼旗下《疯狂动物城》则第一次把城外海选和后台实景搬到漫展现场。阵容亮点之外,本届RED LAND延续“开放世界冒险岛”概念,岛
▸展开看内容
在评估了项目的开发状态与未来方向后,PUBG STUDIOS 认为当前的路线无法带来能让玩家长期享受的游戏体验。因此,官方做出了艰难的决定,决定终止《绝地求生:黑域撤离》的开发。 开发团队对所有参与封闭内测并提供反馈的玩家表示感谢。开发团队表示,会将开发和测试期间获得的经验运用到未来的项目中,并在探索全新玩法时继续倾听玩家意见。
▸展开看内容
Bushiroad Inc.将于2026年10月2日(星期五)起举办“Bushiroad Autumn Sale 2026”,旗下 Nintendo Switch 下载版及 Steam 版游戏将在活动期间以限时优惠价格发售。 Bushiroad旗下游戏的 Nintendo Switch 下载版与 Steam 版现已开启 Autumn Sale,活动期间可享限时优惠价格。本次活动中,《Revue Starlight El Dorado》、《ROAD59 -新时代任侠特区- 摩天楼黑白抗争》、《VIRTUAL GIRL @ WORLD'S END》、《Cardfight!! Vanguard Dear Days 2》等热门作品的下载版均可享受优惠价格。欢迎把握本次机会,畅玩Bushiroad旗下的人气作品。●Bushiroad Autumn Sale 2026 Steam版优惠结束时间:10/9 1:00(UTC+8) Nintendo Switch 下载版优惠结束时间:10/8 22:59(UTC+8) 重点优惠作品⚫ROAD59 -新时代任侠特区- 摩天楼黑白抗争 优惠:较通常价格优惠
▸展开看内容
本期 itch 游戏汇编如期而至,开始阅读,随时试玩,享受时光吧!
▸展开看内容
Focus Entertainment 发行团队与 Afterburner Studios 欣然宣布,于 8 月 25 日科隆游戏展开幕夜直播首次公开的《艾尔塔:逆神》,将于 2027 年 2 月 2 日登陆 PC(Steam)、PlayStation 5 以及 Xbox Series X|S 平台。为让玩家亲身体验这款节奏火爆的动作平台游戏,全平台现已上线免费试玩版,游戏本体同步开启预购。 在免费试玩版中掌握爆发力十足的战斗与流畅的平台跑酷玩法 自 10 月 1 日起,玩家可在全平台(Steam、PlayStation 5、Xbox Series)体验《艾尔塔:逆神》开篇内容的免费试玩版。试玩时长 45 至 60 分钟,玩家将化身武僧战士维克图斯,踏上寻找失落龙神的旅途。这段旅程包含凌厉武术招式、精准平台跳跃与科幻风格探索。在这段开篇冒险的终点,玩家将迎来首场 Boss 战,与 “坎洛亚的选民”展开史诗对决。 现已开启游戏预购随着试玩版上线,游戏已于 2027 年 2 月 2 日正式发售前开启预购。在游戏售价普遍上涨的当下,团队希望兼顾玩家的时间与开支,游戏完整流程时长为 15 至
▸展开看内容
用ImageNet训练encoder
▸展开看内容
🐧👉《再荐茶社》更新通知QQ群:1055561149感谢大家一直以来的赞赏、订阅和支持,欢迎朋友们在评论中和我们一起分享游戏体验!本期我们分享了一些游戏翻译小故事,并且聊了聊我们觉得很优秀、很有趣或者很糟糕的游戏本地化例子。如果你也有遇到过很有趣的翻译,欢迎在评论区分享!
▸展开看内容
草蜢工作室宣布与网易游戏分道扬镳,重新恢复独立。在2021年,网易游戏曾从 GungHo Online Entertainment 手中收购了该工作室。草蜢工作室 CEO 须田刚一在谈到工作室重获独立时表示,工作室于今年2月推出了最新作品《罗密欧是个绝命侠》,并在3月迎来了成立28周年。 在广大玩家的支持下,这部作品获得了极高的评价与良好口碑。工作室对所有喜爱这款游戏以及为游戏发售付出努力的人表示感谢。在完成这部全员参与的项目后,团队在规划未来发展时,产生了为全球玩家带来更多元风格作品的愿望。因此,草蜢工作室选择再次以独立开发者的身份向前发展。据悉,这一决定是在与网易游戏进行充分讨论并获得对方全力配合后做出的。双方将珍惜过去建立的合作关系,并在未来合适的时候期待进一步合作。随着自主发行经验的提升,工作室拥有了更广泛的选择空间。无论游戏采用何种发行方式,草蜢工作室都将坚持打造具有自身特色和更高品质的独特游戏。工作室希望在迈向30周年及更远未来的新征程中,继续得到外界的关注,并对未来创作的作品保持期待与支持。
▸展开看内容
国庆假期和steam秋季特卖一起前来!受秋促影响本周新游戏少了一些,一起来看看吧!
▸展开看内容
本期主播与嘉宾: 脑潘 / 潮玩VR / 空间绘画KudoAlbus / 视界任意门 直播录制时间: 2026-09-30 ----------------------- 主观片面的「元语咒」是专属 VR 玩家的闲聊电台。 以虚拟圆桌的形式在某平台直播,以一刀不剪的态度在这里回放。 ----------------------- 章节: 00:00 开场 02:02 Meta Connect 2026 现场有哪些变化 10:36 发布会之外的观察:平台资源流向 AI、Horizon 与智能眼镜 19:09 Meta VR Glasses 现场体验反馈与讨论,谁会为它买单? 26:03 为了 “看起来像眼镜” 牺牲 FOV 真的值得吗 39:20 内容消费和使用场景的顾虑 50:02 Meta VR 游戏生态的转向,与对 Quest 4 的期待 01:02:32 从整场发布会看 Meta 的重心:AI 服务与智能眼镜 01:10:40 Steam Frame 这边人声鼎沸:涌现海量社区软硬件项目 01:28:26 行业又回到百花齐放:有线与无线的轮回,标准分化带来的开发压力与新机会 01
理科前沿
1 条
▸展开看内容
千禧年难题的突破,10000个Agent最多占了10%的功劳。
文化
1 条
▸展开看内容
凡人终有一死,但从未有过死去的人返回现世,告诉我们那个世界的样子。这恰恰激发了艺术家对于死后世界的想象。从中世纪的死亡之舞,到墨西哥的亡灵狂欢,用不同国家、不同文明的绘画与电影,和你一起拆解令人迷恋的“死亡美学”。
AI 与前沿
31 条
1/9
1
财联社热门102:40
2
3
4
牛客302:40
5
稀土掘金402:40
6
财联社热门502:40
7
Solidot602:40
8
稀土掘金802:40
9
金十数据902:40
10
今日头条902:40
11
V2EX902:40
12
IT之家1102:40
13
14
V2EX1202:40
15
时政军事
10 条
2/9
1
澎湃新闻102:40
2
澎湃新闻202:40
3
参考消息602:40
5
金十数据1102:40
6
联合早报1102:40
7
参考消息1502:40
8
金十数据1702:40
社会与就业
8 条
3/9
1
牛客102:40
2
华尔街见闻102:40
3
华尔街见闻602:40
4
凤凰网1202:40
5
6
7
金十数据2602:40
8
联合早报2702:40
文化
7 条
4/9
1
参考消息802:40
2
知乎902:40
3
澎湃新闻1402:40
4
bilibili 热搜1902:40
5
微博2002:40
论文与开源
6 条
5/9
1
V2EX1102:40
2
稀土掘金1302:40
3
稀土掘金1802:40
4
稀土掘金2102:40
5
6
稀土掘金2902:40
财经商业
5 条
6/9
1
金十数据302:40
2
财联社热门402:40
4
bilibili 热搜1102:40
5
今日头条2402:40
游戏
5 条
7/9
1
Solidot102:40
2
Solidot502:40
3
知乎802:40
4
贴吧1502:40
5
V2EX2802:40
理科前沿
4 条
8/9
1
V2EX502:40
2
3
澎湃新闻1602:40
4
澎湃新闻2002:40
RSS 订阅更新
61 条
AI 与前沿
15 条
▸展开看内容
arXiv:2610.00015v1 Announce Type: new Abstract: Large-language-model agents can propose and execute actions, but proposal, authority, dispatch, verified external effect, and serving promotion are different claims. We present Praxa, an agent harness that represents these states explicitly through deterministic admission, brokered execution, external read-back, reconciliation, and reviewed promotion. We report four evidence lanes. First, an author-run repository-local audit at a pinned revision pa
▸展开看内容
arXiv:2610.00074v1 Announce Type: new Abstract: K-Dense BYOK (bring your own keys) is a free, open-source AI research assistant for scientists in any field that runs on the researcher's own computer. The researcher supplies access to a model of their choice, hosted or running locally, and the application supplies everything else: a place for the work to run, a layer of scientific scaffolding, and a complete record. Each project is an ordinary folder, so the data, the code, the results, and the r
▸展开看内容
arXiv:2610.00529v1 Announce Type: new Abstract: The ontology-based contextual AI evaluation (OB-CAIE) methodology was developed to address a lack of scientific rigor that arises from unclear testing coverage, to balance human expertise and automations, and to address a lack of reproducibility of AI evaluation testing environments. OB-CAIE strengthens the current state of AI evaluations by addressing the first step in the scientific method by clearly defining what will be tested. Two ontologies r
▸展开看内容
arXiv:2610.00531v1 Announce Type: new Abstract: AI-generated content, often called AI slop, is increasingly common everywhere, particularly in academia. Slop in AI-generated scientific papers, however, has more complex patterns that cannot be easily detected by existing token-based AI detectors. Each part of such a paper looks plausible while the scientific reasoning that connects the parts breaks down, which can mislead how readers assess the work. We benchmark these failures as scientific slop
▸展开看内容
arXiv:2610.00636v1 Announce Type: new Abstract: Evaluating AI agents on scientific research tasks is constrained by the time and resources required for the underlying experiments or calculations. In computational materials research, repeating the same expensive simulations across agents and trials can make evaluation impractical. We introduce CompMat-Bench, a benchmark of 94 tasks derived from recently published computational materials studies, each asking agents to complete a step toward achiev
▸展开看内容
arXiv:2610.00654v1 Announce Type: new Abstract: Personalization has long relied on customer data to infer what an individual is likely to value. We call this customer evidence: the customer's historical behavior and preferences. Generative AI extends personalization by allowing providers to supply changing situational information at the moment a response is produced, without encoding every condition in advance. We define this provider-side context as information about what is possible, permitted
▸展开看内容
arXiv:2610.00682v1 Announce Type: new Abstract: Large language models (LLMs) increasingly underpin scientific AI applications that reason over structured knowledge, from biomedical question answering to materials informatics. However, their logical reasoning often falls short, producing factual inaccuracies unacceptable in these settings. Reliable evaluation remains challenging: manual dataset construction scales poorly, and LLM-based generation risks embedding the very flaws it aims to measure
▸展开看内容
arXiv:2610.00791v1 Announce Type: new Abstract: Enterprise information is represented through artifacts shaped by applications, projects, technologies, organizational boundaries, and local requirements. These structures accumulate over time, creating representational complexity that must be maintained by the enterprise and interpreted by information consumers and AI systems. This paper introduces Enterprise Representation Simplification (ERS) as reducing unnecessary representational complexity w
▸展开看内容
arXiv:2610.00797v1 Announce Type: new Abstract: Contextual security defenses prevent AI agents from taking rogue actions by synthesizing a task-specific policy and enforcing it on the agent's tool calls. In multi-step tasks, however, which actions are valid often depends on what the agent has already done and learned. We present Sapien, a policy engine for enforcing stateful contextual policies. A Sapien policy specifies permitted tool-call sequences using a regular expression extended with stat
▸展开看内容
华为 Mate90 系列发布,售价 5999 元起 10 月 1 日,华为 Mate90 系列手机正式发布。华为 Mate90 售价 5999 元起,华为 Mate90 Pro 售价 6999 元起,华为 Mate90 Pro Max 售价 9499 元起,华为 Mate90 Pro Max 典藏版售价 10999 元起,华为 Mate90 RS 非凡大师售价 12999 元起。 影像系统是本次迭代的核心亮点。华为 Mate90 系列全系搭载第三代红枫原色摄像头,标配红外光学镀膜技术,有效提升夜景画面纯净度与通透感。新机新增烟花人像、柔光人像、星芒特效等多元拍摄风格,依托原生原片直出的优质画质,既满足日常随手拍需求,也为后期创作预留充足空间。同时,全新华为光变巨炮睿影 Z10 专业影像模块正式登场,区别于传统增距镜设备,其内置 1 英寸大底传感器,支持 1-10 倍连续光学变焦,大幅拓展移动端专业拍摄边界。(来源:36 氪) 苹果首款 OLED 触屏 MacBook Pro 或最快本月发布 苹果首款配备 OLED 显示屏的 MacBook Pro 正在逐渐接近发布。彭博社记者马克·古尔
▸展开看内容
我的“声音”到底归谁?我的“声音”又到底值多少钱?当AI开始轻而易举的模仿我们的声音肆意表达,许多过去从未出现的问题逐一浮现。知名声优、演员、主持人津田健次郎今日起诉tiktok平台引发争议,而在播客圈儿,也不断有我们熟悉的主播声音出现在互联网的视频中,所以我们决定从“声音的价值”这个角度切入,从广播剧的诞生,到声优行业的出现,再到播客时代,被录下的声音如何定价、授权?百年变化背后,藏着和你有关的版权难题,但也似乎是在给ai时代的困境指明了一条出路。【时间轴】02:01 声音背后的争议:从津田健次郎在TIKTOK上的诉讼案例聊起 10:40 探索百年人类声音记录的科技历程… 21:20 日本近现代文学中的“言文一致”运动:平民百姓的文化需求与创作潮流 32:01 声优行业的起点:1925年广播剧研究与日本声优史的关系 42:43 动画片制作背后的艰辛与收益分配:手冢治虫与后续动画行业的博弈 53:24 演员罢工引发的连锁反应:电视台滚动更新受阻,配音行业的重要性逐渐被重视 01:04:03 日本排优联合提出动画重播费用问题,动画配音演员的待遇能否改善? 01:14:44 声音的版权争议
▸展开看内容
一个工程师写完数据处理代码,点下提交,然后去喝了杯咖啡。十几分钟甚至几十分钟之后,他才回到屏幕前看结果、改代码、再提交。这条「等待— 反馈」的空隙,过去一直是数据平台默认的节奏。 Agent 没有这个空隙。它拿到一次结果,立刻验证,紧接着并发地去试第二种、第三种方案,飞快拿到反馈,再发起下一轮调用。 人用平台,发过来的是一个相对确定的任务,算完、给结果、走人;Agent 用平台,会不停地换着法子去完成你交代的那个目标。压力就这样被放大了—— 管控的压力和隐形计算的压力,会十倍、百倍地往上翻。 这正在成为 2026 年云计算需要面对的一种新负载。 AI 正在深入「把模型用起来」的阶段。模型一旦长成 Agent,它就不再是一个躺在那里等你调用的接口,而更像一个会自己找数据、自己拆任务、自己并发试错、自己一直跑下去的数字员工。它制造出来的负载,和过去十几年云平台习惯承载的那种负载,完全是两回事。 阿里云智能集团首席技术官李飞飞在云栖大会上把智能体落地拆成三个词:Model 决定智能的上限,Context 决定 Agent 读不读得懂你的业务,Harness 决定它能不能真正调起工具、把事做完
▸展开看内容
苹果即将发布「智能家居中枢」新品 据彭博社报道,苹果公司计划于 10 月 13 日推进其拖延已久的智能家居市场布局,这是该公司在新任 CEO 约翰·特努斯领导下的一次关键产品扩张。 据知情人士透露,该战略的核心是一款代号为 J490 的智能家居中枢设备。该设备将采用大约 6 英寸的方形显示屏设计,提供可安装在墙上或放置在台面上的版本。 按照苹果的设想,用户会在家中多处放置多台这种显示屏。它们可用于控制恒温器、门锁和其他联网产品,也可用于视频通话、对讲式沟通、播放音乐,以及查看存储在苹果 iCloud 服务中的照片幻灯片。 苹果还把这台设备视为厨房里的通信和信息中枢,用户可以在做饭时通过语音查询菜谱、观看视频或拨打 FaceTime 通话。 同时,报道称苹果也计划宣布 HomePod mini 自 2020 年首次亮相以来的首次更新,以及自 2022 年以来的首款新电视机顶盒。(来源:凤凰网科技) 谷歌 Gemini 4 Argon 登场,长周期代码工程测试超 Opus 5.5 当地时间 9 月 30 日,谷歌发布 Gemini 4 Argon,称其为公司迄今最先进的 AI 模型,但目前
论文与开源
8 条
▸展开看内容
arXiv:2610.00012v1 Announce Type: new Abstract: LLM agents increasingly act through modular systems, such as order, payment, inventory, and shipment services, where actions in one module change which transitions are valid in another. Standard world models usually fit observational traces, but this is not the quantity needed for intervention-time planning: a trace may show that payment precedes shipment without identifying whether payment authorizes shipment, inventory mediates the effect, or a h
▸展开看内容
arXiv:2610.00025v1 Announce Type: new Abstract: Agent harnesses increasingly want to run small language models (SLMs) on the microtasks around a frontier large language model (LLM) planner: auto-approving shell commands, writing memory, selecting tools, ranking past turns. We ask whether off-the-shelf SLMs meet practitioner-defined thresholds and, when they fail, why, and whether quantization changes the answer. We build a benchmark of 4 such microtasks with fixed prompts and automatic metrics
▸展开看内容
arXiv:2610.00313v1 Announce Type: new Abstract: Scientific coding agents receive equations, boundary conditions, and output requirements in writing, then must assess the programs they revise. Rules to Tools (R2T) supplies prepared executable checks of public scientific requirements. Matched SciCode repair groups share written checks, starting programs, model, and budgets; the tool group receives a callable implementation. Across two task-ID cohorts, complete repair is 26/30 with text and 29/30 w
▸展开看内容
arXiv:2610.00349v1 Announce Type: new Abstract: Resource limits are becoming an authorization boundary for AI agents that delegate work across concurrent and failure-prone workers. Parent-child allocation constraints, affine objects, and distributed escrow do not by themselves prevent overspend when replies are lost, effects complete after timeout, messages repeat, branches partition, or DAG joins alias one lineage. We formalize fault-tolerant budget conservation for distributed multi-agent dele
▸展开看内容
arXiv:2610.00366v1 Announce Type: new Abstract: A persistent agent must decide both what to retain as information arrives and what to surface once a query appears, yet memory evaluations can confound these decisions by comparing methods that differ in both retention and selection. We build a streaming-recall benchmark crossing retention and selection rules and evaluate every condition on the same 300 seeded episodes. Holding access fixed, query-aware selection improves required-fact recall by 15
▸展开看内容
arXiv:2610.00372v1 Announce Type: new Abstract: Large language model agents rely on external harnesses to pass information between the model and its environment and to recover from execution errors. Yet recovery is usually judged only by average task success. This hides an important tension. The same operation can rescue a failing trajectory or disrupt one that would otherwise succeed. We frame recovery as a causal decision problem. Starting from the same execution state, we compare what happens
游戏
10 条
▸展开看内容
10月2日,小红书开放世界冒险岛RED LAND正式开幕,上海杨浦复兴岛再次变身”痛岛”。早上不到10点,船台公园入口处已经排起了长队。这是RED LAND连续第二年落地复兴岛,活动将持续至10月6日。相较去年,本届RED LAND规模全面升级:场地从8万平方米扩展至15万平方米,超过160个游戏及动漫IP登岛,双双翻倍。据官方介绍,本届五轮售票平均每轮20.6秒全部售罄,这场持续五天的同好狂欢,也成为今年国庆假期最受关注的青年潮流文化活动之一。 160+个人气IP登岛,打造2.5次元“沉浸式开放世界”走进复兴岛,巨幅人气角色气模簇拥着机械船台,身着各式cos装扮的玩家与岛民NPC穿梭往来,在这里,青年潮流文化与工业遗存交相辉映。本届RED LAND阵容不仅集结了160+人气IP,也集齐了多个“首次”。米哈游旗下四款作品首次线下同框,其生活模拟游戏新作《星布谷地》迎来国内首次线下亮相;复联、X战警、神奇四侠与黑豹五座英雄巨像在《复仇者联盟5》展位全球首亮相,迪士尼旗下《疯狂动物城》则第一次把城外海选和后台实景搬到漫展现场。阵容亮点之外,本届RED LAND延续“开放世界冒险岛”概念,岛
▸展开看内容
在评估了项目的开发状态与未来方向后,PUBG STUDIOS 认为当前的路线无法带来能让玩家长期享受的游戏体验。因此,官方做出了艰难的决定,决定终止《绝地求生:黑域撤离》的开发。 开发团队对所有参与封闭内测并提供反馈的玩家表示感谢。开发团队表示,会将开发和测试期间获得的经验运用到未来的项目中,并在探索全新玩法时继续倾听玩家意见。
▸展开看内容
Bushiroad Inc.将于2026年10月2日(星期五)起举办“Bushiroad Autumn Sale 2026”,旗下 Nintendo Switch 下载版及 Steam 版游戏将在活动期间以限时优惠价格发售。 Bushiroad旗下游戏的 Nintendo Switch 下载版与 Steam 版现已开启 Autumn Sale,活动期间可享限时优惠价格。本次活动中,《Revue Starlight El Dorado》、《ROAD59 -新时代任侠特区- 摩天楼黑白抗争》、《VIRTUAL GIRL @ WORLD'S END》、《Cardfight!! Vanguard Dear Days 2》等热门作品的下载版均可享受优惠价格。欢迎把握本次机会,畅玩Bushiroad旗下的人气作品。●Bushiroad Autumn Sale 2026 Steam版优惠结束时间:10/9 1:00(UTC+8) Nintendo Switch 下载版优惠结束时间:10/8 22:59(UTC+8) 重点优惠作品⚫ROAD59 -新时代任侠特区- 摩天楼黑白抗争 优惠:较通常价格优惠
▸展开看内容
本期 itch 游戏汇编如期而至,开始阅读,随时试玩,享受时光吧!
▸展开看内容
Focus Entertainment 发行团队与 Afterburner Studios 欣然宣布,于 8 月 25 日科隆游戏展开幕夜直播首次公开的《艾尔塔:逆神》,将于 2027 年 2 月 2 日登陆 PC(Steam)、PlayStation 5 以及 Xbox Series X|S 平台。为让玩家亲身体验这款节奏火爆的动作平台游戏,全平台现已上线免费试玩版,游戏本体同步开启预购。 在免费试玩版中掌握爆发力十足的战斗与流畅的平台跑酷玩法 自 10 月 1 日起,玩家可在全平台(Steam、PlayStation 5、Xbox Series)体验《艾尔塔:逆神》开篇内容的免费试玩版。试玩时长 45 至 60 分钟,玩家将化身武僧战士维克图斯,踏上寻找失落龙神的旅途。这段旅程包含凌厉武术招式、精准平台跳跃与科幻风格探索。在这段开篇冒险的终点,玩家将迎来首场 Boss 战,与 “坎洛亚的选民”展开史诗对决。 现已开启游戏预购随着试玩版上线,游戏已于 2027 年 2 月 2 日正式发售前开启预购。在游戏售价普遍上涨的当下,团队希望兼顾玩家的时间与开支,游戏完整流程时长为 15 至
▸展开看内容
用ImageNet训练encoder
▸展开看内容
🐧👉《再荐茶社》更新通知QQ群:1055561149感谢大家一直以来的赞赏、订阅和支持,欢迎朋友们在评论中和我们一起分享游戏体验!本期我们分享了一些游戏翻译小故事,并且聊了聊我们觉得很优秀、很有趣或者很糟糕的游戏本地化例子。如果你也有遇到过很有趣的翻译,欢迎在评论区分享!
▸展开看内容
草蜢工作室宣布与网易游戏分道扬镳,重新恢复独立。在2021年,网易游戏曾从 GungHo Online Entertainment 手中收购了该工作室。草蜢工作室 CEO 须田刚一在谈到工作室重获独立时表示,工作室于今年2月推出了最新作品《罗密欧是个绝命侠》,并在3月迎来了成立28周年。 在广大玩家的支持下,这部作品获得了极高的评价与良好口碑。工作室对所有喜爱这款游戏以及为游戏发售付出努力的人表示感谢。在完成这部全员参与的项目后,团队在规划未来发展时,产生了为全球玩家带来更多元风格作品的愿望。因此,草蜢工作室选择再次以独立开发者的身份向前发展。据悉,这一决定是在与网易游戏进行充分讨论并获得对方全力配合后做出的。双方将珍惜过去建立的合作关系,并在未来合适的时候期待进一步合作。随着自主发行经验的提升,工作室拥有了更广泛的选择空间。无论游戏采用何种发行方式,草蜢工作室都将坚持打造具有自身特色和更高品质的独特游戏。工作室希望在迈向30周年及更远未来的新征程中,继续得到外界的关注,并对未来创作的作品保持期待与支持。
▸展开看内容
国庆假期和steam秋季特卖一起前来!受秋促影响本周新游戏少了一些,一起来看看吧!
▸展开看内容
本期主播与嘉宾: 脑潘 / 潮玩VR / 空间绘画KudoAlbus / 视界任意门 直播录制时间: 2026-09-30 ----------------------- 主观片面的「元语咒」是专属 VR 玩家的闲聊电台。 以虚拟圆桌的形式在某平台直播,以一刀不剪的态度在这里回放。 ----------------------- 章节: 00:00 开场 02:02 Meta Connect 2026 现场有哪些变化 10:36 发布会之外的观察:平台资源流向 AI、Horizon 与智能眼镜 19:09 Meta VR Glasses 现场体验反馈与讨论,谁会为它买单? 26:03 为了 “看起来像眼镜” 牺牲 FOV 真的值得吗 39:20 内容消费和使用场景的顾虑 50:02 Meta VR 游戏生态的转向,与对 Quest 4 的期待 01:02:32 从整场发布会看 Meta 的重心:AI 服务与智能眼镜 01:10:40 Steam Frame 这边人声鼎沸:涌现海量社区软硬件项目 01:28:26 行业又回到百花齐放:有线与无线的轮回,标准分化带来的开发压力与新机会 01
理科前沿
1 条
▸展开看内容
千禧年难题的突破,10000个Agent最多占了10%的功劳。
文化
1 条
▸展开看内容
凡人终有一死,但从未有过死去的人返回现世,告诉我们那个世界的样子。这恰恰激发了艺术家对于死后世界的想象。从中世纪的死亡之舞,到墨西哥的亡灵狂欢,用不同国家、不同文明的绘画与电影,和你一起拆解令人迷恋的“死亡美学”。