#The

智能AI None

嵌入模型路由的政策遗憾：低级别专家的上下文强盗

14929v1 Announce Type: new Abstract: Modern recommendation systems increasingly rely on dynamically routing diverse quer

and the models

2026-06-16 Yan Dai, Negin Golrezaei, Patrick Jaillet

智能AI None

{\alpha}-公平保险定价：公平连续体

14898v1 Announce Type: new Abstract: Fairness in insurance pricing remains a long-standing and deeply debated puzzle。

and fairness insurance

2026-06-16 Tianhe Zhang, Xiguang Liu, Peng Shi

智能AI None

GRAPE：用于紧凑对抗鲁棒性的引导参数空间演化

14865v1 Announce Type: new Abstract: Adversarial Training (AT) improves neural network robustness, but most methods trai

the parameter space

2026-06-16 Zhiyuan Ye (University of Science and Technology of China), Xiangyu Zhou (China Mobile), Ji Qi (China Mobile), Hao Zhang (University of Science and Technology of China), Yi Zhou (China Mobile)

智能AI None

QPILOTS：针对流量策略的高效测试时 Q 转向

14801v1 Announce Type: new Abstract: Flow-matching and diffusion policies are expressive action generators, but optimizi

the critic QPILOTS

2026-06-16 Yifan Ruan, Chenyang Cao, Andreas Burger, Ali Pesaranghader, Kaveh Kamali, Jaehong Kim, Nandita Vijaykumar, Alan Aspuru-Guzik, Igor Gilitschenski, Nicholas Rhinehart

智能AI None

指标匹配：评估法学硕士法官可靠性的子集选择方法

15029v1 Announce Type: new Abstract: LLM judges are used to reduce the need for costly human labor in evaluating open-en

for reliability the

2026-06-16 Alyssa Unell, Natalie Dullerud, Naomi Boneh, Meena Jagadeesan, Tatsu Hashimoto, Nigam Shah, Sanmi Koyejo

智能AI None

AI Engram：寻找人工智能中的记忆痕迹

arXiv:2606.14997v1 Announce Type: new Abstract: Memory formation is fundamental to intelligence, yet whether deep neural

and memory engrams

2026-06-16 Jea Kwon, Dong-Kyum Kim, Jiwon Kim, Yonghyun Kim, Woong Kook, Meeyoung Cha

智能AI None

PrologMCP：LLM 代理的标准化 Prolog 工具接口

arXiv:2606.14935v1 Announce Type: new Abstract: Frontier reasoning-tuned language models still fail on deductive tasks a

the reasoning and

2026-06-16 Agnieszka Mensfelt, Adarsh Prabhakaran, Adrian Haret, Vince Trencsenyi, Kostas Stathis

智能AI None

人工智能代理之间的信任：测量形成、破坏和恢复，对治理多代理系统的影响

14923v1 Announce Type: new Abstract: As language-model agents increasingly work in teams, each agent must decide how muc

trust the and

2026-06-16 Yujiao Chen

智能AI None

良好解释的定义以及解释法学硕士成果的挑战

14838v1 Announce Type: new Abstract: How to define a good explanation is a long-standing philosophical debate which has

good the explanations

2026-06-16 Louis Mahon, Elliot Ford, Callum Hackett

智能AI None

为什么韩国人如此热爱人工智能？

This story originally appeared in The Algorithm, our weekly newsletter on AI。To get stories like this in your inbox firs

This story originally

2026-06-16 Michelle Kim

智能AI None

腾讯投了林俊旸1.4亿！前千问负责人创业，估值135亿

新智元报道 【新智元导读】 刚离职就狂揽135亿。前阿里千问领军人林俊旸闪电创业，腾讯、红杉疯狂抢投。这位AI新贵，究竟押中了什么神级赛道。就在昨天，外媒The Information爆料—— 前阿里巴巴千问大模型负责人林俊旸创办的AI实验

具身智能 世界模型 The

2026-06-16 新智元

数码硬件 None

GNU Linux-libre 7.1-gnu 发布，带有更多驱动程序调试，对 i486 删除不满意

Following yesterday's release of the upstream Linux 7。1 kernel release, GNU Linux-libre 7。

the release Linux

2026-06-15 Michael Larabel

数码硬件 None

Linux 7.2 更好地交流文件系统案例折叠以帮助 Windows NFS 等

Newly-merged code for the in-development Linux 7。

the case Linux

2026-06-15 Michael Larabel

数码硬件 None

Linux 7.2 引入 Rust Zerocopy 库以消除更多“不安全”代码

Miguel Ojeda already mailed in the many Rust code changes for the in-development Linux 7。This is quite a big Rust code w

the Rust code

2026-06-15 Michael Larabel

数码硬件 None

Linux 7.2 合并缓存感知调度以增强现代 Intel 和 AMD CPU

The scheduler updates were merged this morning for the Linux 7。2 kernel and it's exciting。Cache Aware Scheduling has fin

merged for and

2026-06-15 Michael Larabel

数码硬件 None

令人惊讶的是，读取 /proc/filesystems 的频率非常高，现在速度提高了 444%

Reading /proc/filesystems for obtaining a list of file-systems supported by the running kernel is done frequently on Lin

the proc filesystems

2026-06-15 Michael Larabel

数码硬件 None

Firefox 152 现已推出，默认内置 JPEG-XL 支持，现代化设置 UI

The Firefox 152。0 release binaries are now available ahead of tomorrow's official unveiling。

now Firefox release

2026-06-15 Michael Larabel

数码硬件 None

俄罗斯垃圾邮件和脏话现在困扰着 Arch Linux AUR

After days of dealing with 1,500+ packages in the Arch Linux AUR containing malware, the latest headache in the Arch Lin

the Arch Linux

2026-06-15 Michael Larabel

开发者生态 None

伊罗1.0

It's a simple idea really, and it's the right abstraction for the future of the internet。IP addresses can break, without

the can and

2026-06-15 chadfowler

开发者生态 None

Show HN：利用Slack的视频嵌入实现E2EE通信

Versión en Español aquí ; gh:v1ctorio/e2ee-slack Introduction Some time ago, while exploring Slack’s Block Kit reference

the and video

2026-06-15 victorio