Skip to main content

XMT

短闻

信流 · 上滑连读 · 来源可核

今日 稍后 搜索 RSS
1 / 24
Aggregate arXiv cs.AI 人工智能 45″

Dual-Cache Latent Space Communication between Heterogeneous Language Models

arXiv:2608.…

  • 20617v1 Announce Type: new Abstract: Multi-agent LLM systems split wor…
  • They usually communicate by exchanging text, which puts autoregressive…
  • Recent latent protocols instead translate the sharer's key-value (KV) …

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Evaluating Skills, Not Just Agents: Agentic Continuous Evaluation of Skills

arXiv:2608.…

  • 20614v1 Announce Type: new Abstract: Enterprise agent programs are mov…
  • Current gates often scan these artifacts for structure, style, and sec…
  • We present ACES (Agentic Continuous Evaluation of Skills), a repositor…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Difficulty-Aware Semantic-ID Optimization for Generative Recommendation

arXiv:2608.…

  • 20611v1 Announce Type: new Abstract: Semantic-ID-based generative reco…
  • A common recipe is SFT followed by GRPO, yet vanilla GRPO is poorly ma…
  • Under the frozen SFT checkpoint, the exact target is absent from the f…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

FlavourBench: Ranking Frontier Language Models with Executable Culinary Ground Truth

arXiv:2608.20574v1 Announce Type: new Abstract: Open-ended language-model benchmarks usually inherit a judge: a human preference panel, another model, or a brittle exact-match key.…

  • We introduce FlavourBench, an automated benchmark in which a versioned…
  • Each task presents eight ingredients and asks for a three-ingredient p…
  • We evaluate 27 frontier endpoints on an identical 534-task core spanni…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Open-Weight Masked Introspection: Measuring What Language Models Can Report About Their Own Computation

arXiv:2608.20569v1 Announce Type: new Abstract: Are frontier models able to introspect about their internal states?…

  • Recent work suggests that under certain conditions a complex enough mo…
  • We tested that claim on eight open-weight models from seven families a…
  • To test it we built Open-Weight Masked Introspection (OWMI), a framewo…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate 数英网 时尚 45″

文章频道 - 通用AI助手,难逃互联网老路?

AI助手会推荐,不等于会做生意。文|刘诗雨;编辑|孙静;来源:降噪NoNoiseAI助手终于开始认真谈钱了。这个夏天,国内头部AI助手陆续把收费摆上台面。千问开始尝试会员服务,豆包则两头下注:一边推出专业版,直接从用户钱包里收钱;另一边盯上酒店等本地生活订单,准备从交易结果里抽佣。…

  • 过去几年,大厂做AI助手的逻辑统一:先卷用户规模,赚钱以后再说。
  • 当行业还在争论谁是「中国ChatGPT」时,用户规模几乎是首要指标。
  • 仿佛谁先成为几亿人手机里的AI入口,盈利模式便会自然长出来。

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate 爱范儿 人工智能 45″

折叠 iPhone 最全爆料汇总:用过的都说好,但有两个大缺点

今年是折叠屏诞生的第七个年头,苹果终于要交出他的第一份答卷。彭博社记者 Mark Gurman 给出了迄今为止最接近「上手体验」的爆料。…

  • 按照 Gurman 的说法,他已经与数名实际使用过折叠 iPhone 的人士交流,几名体验者给出的反馈颇为一致: 这台手机放进口袋里的感觉不…
  • Gurman 提到,折叠 iPhone 的影像表现也可圈可点——这意味着苹果并不会像 iPhone Air 那样走极端,放弃 iPhone …
  • 当然,这台折叠屏的售价,可能是更大的问题——目前来看,折叠 iPhone 售价估计会超过 2000 美元,折合人民币超过一万五。

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate 雷锋网 人工智能 45″

阿里视频大模型Wan3.0正式上线,行业评价“稳定、真实、有质感”

8月24日,阿里巴巴视频生成大模型Wan3.0正式上线。该模型在生成时长、万能创作、全能参考以及真实世界还原等维度全面升级,单次可生成30秒视频,并首次支持doc、xls、ppt、pdf、md等文档格式输入,力求准确还原真实世界。…

  • 公测以来,见证Wan模型家族历次更新的创作者表示,Wan3.
  • 从企业用户体验来看,Wan3.
  • 0收到的反馈高度一致,“稳定、真实、有质感”是被反复提到的三个关键词。

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate 雷锋网 人工智能 45″

王田苗三问具身大牛:大模型还在「拨号上网」,机器人靠什么拿下真实订单?| WRC 2026

当狂热的资本逐渐冷静,具身智能行业必须面对一场从“展示Demo”到“真实产线”的生死大考。编辑丨岑峰 “如果大模型现在还处在‘拨号上网’的时代,那谁能告诉我,制约具身智能规模化的致命卡点到底是什么?…

  • 大模型解决物理世界问题的巅峰时刻,究竟何时到来?
  • ”在 2026 世界机器人大会(WRC)的一场重磅圆桌论坛上,北京航空航天大学教授、中国机器人学术泰斗王田苗,向台下的五位具身智能头部企业 …
  • 过去两年,大模型的狂飙将具身智能推上了风口浪尖。

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate 雷锋网 人工智能 45″

WRC 2026|原生全模态世界模型:从模拟世界到交互世界

8月19日,世界机器人大会(WRC 2026)期间,由跨维智能主办、《财经》杂志协办的“物理AI引领者论坛暨2026《财经》中国AI 100榜单启动仪式”在北京举行。论坛以“物理AI进行时”为主题,汇聚具身智能领域的顶尖学者、企业创始人和一线技术负责人,共同探讨通用物理智能的技术演进与产业路径。…

  • 作为原生全模态世界模型代表企业,智象未来(HiDream.
  • ai)创始人兼CEO梅涛发表了题为《原生全模态世界模型:从“模拟世界”到“交互世界”》的主旨演讲。
  • 他从AI大模型的技术演进趋势出发,系统阐述了大模型从多模态走向原生全模态、从模拟世界走向交互世界的核心逻辑,为Physical AI的底层能…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Consilience: Conformally Calibrated Communication Control for Hidden-Profile Multi-Agent Reasoning

arXiv:2608.…

  • 20564v1 Announce Type: new Abstract: Multi-agent LLM systems can impro…
  • Existing protocols, including fixed schedules, round-robin exchange, a…
  • We propose Consilience, an inference-time orchestration framework that…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Volumetric Radiology AI in the Era of Multimodal Large Language Models

arXiv:2608.…

  • 20549v1 Announce Type: new Abstract: Advances in multimodal large lang…
  • Volumetric radiology, however, presents a fundamental representational…
  • Reliable volumetric radiology AI therefore requires representations th…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

FL-MAESTRO: Multi-Agent LLM Orchestration for Resource-Constrained Federated Learning

arXiv:2608.…

  • 20518v1 Announce Type: new Abstract: In Federated Learning (FL), the c…
  • Each round, the server must commit three coupled decisions, namely the…
  • Recent agentic systems have begun bringing large language models (LLM)…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

A Temporal Planning Approach for Intelligent Flood Response

arXiv:2608.…

  • 20510v1 Announce Type: new Abstract: Effective response to multiple, s…
  • Automated planning provides a foundation for addressing this challenge…
  • This work presents an intelligent flood-response framework that exploi…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Lost in Translation: How Universal Ethical Values Fail to Translate Across Global Contexts

arXiv:2608.…

  • 20490v1 Announce Type: new Abstract: AI ethics frameworks treat values…
  • We examined how 14 experts across 10 countries made sense of AI in pra…
  • We found that AI deployment is characterized by structurally unequal c…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate Hypebeast 时尚 45″

Two Souls Bound by Impulse: The Cinematic Poetry of Valentino’s “Vendetta”

SummaryRomain Gavras directs Dakota Johnson and Alexander Skarsgård in a 35mm campaign film for Valentino Beauty's newest fragrance, Vendetta.…

  • Vendetta is made up of two scents: the tuberose-driven "Vendetta Donna…
  • "Valentino Beauty unveils a magnetic return to passion with its new pr…
  • Directed by Romain Gavras and photographed by Inez & Vinoodh, the cine…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate Hypebeast 时尚 45″

Xbox and IKEA Tease Upcoming Gaming Furniture Collection

SummaryXbox and IKEA have officially begun teasing a collaborative furniture and household accessories collectionEarly teasers preview Xbox-inspired decor including a green D-pad cushion and an analogue stick stoolThe co-designed range is scheduled to debut at the Xbox booth during GamescomXbox and IKEA have officially begun teasing an upcoming collaborative gaming furniture and household accessories collection.…

  • Social media teasers from both brands preview an upcoming crossover be…
  • Early teaser videos on Instagram showcase distinct Xbox-inspired home …
  • Highlight items include a bright green cushion shaped like an Xbox con…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate WWD 时尚 10″

Gap Inc.’s Fashion Week ‘Encore’ Takeover

The company’s Encore loyalty program was launched last February with richer perks and points than the previous one.

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate The Block 金融科技 9″

Bitcoin records largest weekly dollar gain in history; Strive CEO predicts ‘strongest’ cycle ever

Bitcoin records largest weekly dollar gain in history; Strive CEO predicts ‘strongest’ cycle ever

  • Bitcoin records largest weekly dollar gain in history; Strive CEO predicts ‘strongest’ cycle ever

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate STAT News 生物医药 45″

Americans bought 12% less soda under new SNAP restrictions, study says

New restrictions on the use of food benefits to purchase soda and candy have been one of the biggest changes enacted by the Make America Healthy Again movement.…

  • But a key question has been how profoundly the bans would affect publi…
  • Now a new study shows that soda purchases did indeed fall by about 12%…
  • The study, which was published by the National Bureau of Economic Rese…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Terminal Agents: A Survey of AI Agents in Command-Line Environments

arXiv:2608.…

  • 20485v1 Announce Type: new Abstract: Large language model agents incre…
  • We regard terminal agents as systems whose dominant progress-bearing a…
  • Using terminal-mediated execution as an organizing lens, this survey e…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

STCO: Conditional Neural Operators for Time-Dependent PDEs

arXiv:2608.…

  • 20477v1 Announce Type: new Abstract: Neural operators have emerged as …
  • For control or optimization, however, body motion, inflow, or forcing …
  • We introduce the Spatiotemporal Conditional Operator (STCO) for prescr…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Who Delegates to AI? Evidence from 53,000 Agent Configurations

arXiv:2608.…

  • 20425v1 Announce Type: new Abstract: A growing literature measures how…
  • We propose a new layer of exposure, delegated exposure, which records …
  • We operationalize it as the Agentic Adoption Index (AAI), which measur…

RSS 官方收录 · 可信分层展示

详情 原文 分享图
Aggregate arXiv cs.AI 人工智能 45″

Categorical AI phenomenology: A first-person approach

arXiv:2608.…

  • 20420v1 Announce Type: new Abstract: This paper develops a phenomenolo…
  • We shift the methodological focus to first-person structures, modeled …
  • In this framework, Q-networks are conceptualized as relational interfa…

RSS 官方收录 · 可信分层展示

详情 原文 分享图