1
2
Show HN: Capsule – E2E encrypted file sharing
端到端加密、无需注册的命令行文件分享，传文件瞬间生成短链和二维码。
Capsule started as a basic HTTP server to teach myself Rust. Then I kept going, and it turned into the first full-stack thing I've fully shipped: a Ru…
3
宇树科技陈立：机器人目前没有大面积应用，主要原因是具身智能大模型不成熟
宇树科技陈立直言机器人应用瓶颈：具身智能大模型不成熟，统一端到端模型是未来关键，还透露了公司销量、IPO等最新动向。
IT之家 7 月 4 日消息，据新浪财经报道，亚布力中国企业家论坛第十二届创新年会于 7 月 3 日-4 日在上海召开。宇树科技股份有限公司联合创始人陈立表示， 机器人目前其实还没有大面积应用 ， 主要原因还是具身智能大模型没有成熟 。 他首先认为，未来 2-5 年首先需要有一个统一的端到端机器人大…
4
DA-Studio: An Agentic System for End-to-End Data Analysis
VLDB 2026 Demo论文，提出端到端数据分析智能体系统DA-Studio，打通数据获取到洞察全流程。
arXiv:2606.31423v1 Announce Type: cross Abstract: Real-world data analysis is a multi-step process over heterogeneous inputs rather than merely produc…
5
5ting at SemEval-2026 Task 8: Strong End-to-End Multi-Turn RAG via LLM-Based Reranking and Faithfulness Control
首次公开多轮RAG端到端强基线：用LLM重排序+忠诚度控制，在SemEval-2026上验证有效性。
arXiv:2606.28737v1 Announce Type: new Abstract: We introduce 5ting, our system for the SemEval2026 Task 8 (MTRAGEval), which evaluates multi-turn Retr…
6
地平线发布辅助驾驶系统 HSD V2.0：世界模型 + 端到端强化学习，奇瑞 iCAR V27 明日起开放 OTA 升级
IT之家 6 月 29 日消息，地平线今日发布全场景辅助驾驶系统 HSD V2.0 版本，基于世界模型 + 端到端强化学习。 IT之家获悉，地平线 HSD V2.0 即将开启车企推送， 其中 iCAR V27 作为首批搭载车型，将于 6 月 30 日起分批次向车主用户开放 OTA 升级 。更多合作车…
7
追赶FSD V14，理想在补哪些课？｜最前线
理想追赶FSD V14，解析自动驾驶端到端范式下VLA与世界模型整合的关键课业
过去几年，智能驾驶行业的竞争重心经历了几次明显变化。 最早比的是硬件：激光雷达要不要上、摄像头装几个、算力做到多少 TOPS；随后进入大模型时代，竞争开始转向端到端、VLA（Vision-Language-Action）、World Model（世界模型）等路线。 到了今天，越来越多公司发现，仅仅拥…
8
百度开源 Unlimited OCR 模型：基于 DeepSeek OCR，解析文档告别 AI 越生成越慢
百度开源端到端OCR模型，总参30亿仅激活5亿，有效解决文档解析越生成越慢问题
IT之家 6 月 25 日消息，百度于 6 月 22 日开源推出 Unlimited OCR 模型，总参数量 30 亿，推理时仅激活 5 亿参数 ，目标解决在解析长文档时，端到端 OCR 模型越生成越慢的问题。 IT之家注：端到端 OCR 模型是统一神经网络架构系统，融合检测图像中的文本和字符识别，…
9
Green unit tests don't mean you can go live
单元测试全绿也不能轻易上线，端到端真实验证才是关键。
We sell authentication, which makes our scariest question a simple one: when a customer wires their app to us and switches everything on, does it actu…
10
Training the Orchestrator: A Supervised Approach to End-to-End PDDL Planning with LLM Agents
LLM代理结合PDDL规划，端到端有监督训练编排器，突破传统规划瓶颈。
arXiv:2606.21740v1 Announce Type: new Abstract: Translating natural-language planning intent into verified plans is a longstanding challenge: people c…
11
Joint Speech and Text Training for LLM-Based End-to-End Spoken Dialogue State Tracking
端到端口语对话状态追踪新突破：联合语音与文本训练，显著提升LLM在语音场景下的跟踪准确率。
arXiv:2511.22503v2 Announce Type: replace Abstract: End-to-end spoken dialogue state tracking (DST) is made difficult by the tandem of having to handl…
12
Unlimited OCR Works
端到端OCR模型借助大语言模型解码，解锁更强识别能力，DeepSeek OCR引领新突破
arXiv:2606.23050v1 Announce Type: cross Abstract: Recently, end-to-end OCR models, exemplified by DeepSeek OCR, have once again thrust OCR into the sp…
13
Signal’s Meredith Whittaker wants you to remember that AI chatbots ‘are not your friends’
看完这篇文章，你会意识到AI聊天机器人不是真正的朋友——它是一个帮你清醒认识AI隐私风险的意识提升工具，而非实际软件。
"These are not your friends. These are not conscious beings. These are not sentient interlocutors.”
14
15
Leveraging Audio-LLMs to Filter Speech-to-Speech Training Data
利用Audio-LLMs智能过滤低质量数据，大幅提升语音到语音翻译模型的训练效率与效果，INTERSPEECH 2026录用论文。
arXiv:2606.13507v1 Announce Type: new Abstract: Large-scale mined corpora provide abundant training data for end-to-end speech-to-speech translation (…
16
Message Franking: Reporting Abuse Without Breaking Encryption
一种叫“消息弗兰克”的机制，让你在端到端加密聊天中安全举报滥用，不破加密。
Here's a problem that sounds impossible. A messaging app can't read your end-to-end encrypted messages — that's the whole point. But when someone rece…
17
A python terminal writer with git, github e2ee sync, hardware bound keys and temporal search
一款Python终端笔记工具，支持Git/GitHub端到端加密同步、硬件密钥绑定与时间搜索，保护隐私又高效。
📓 Terminal Notes sys-ronin / terminal-notes Terminal notes - Git, encryption, free E2EE sync, cognitive UI 📓 Terminal Notes Encrypted terminal-based w…
18
EPS3D: End-to-End Feed-Forward 3D Panoptic Segmentation
ICML 2026 最新研究，端到端前馈架构实现 3D 全景分割，代码已开源。
arXiv:2606.08980v1 Announce Type: new Abstract: This paper introduces EPS3D, a new end-to-end feed-forward framework for open-vocabulary 3D panoptic s…
19
From Human Guidance to Autonomy: Agent Skill System for End-to-End LLM Deployment on Spatial NPUs
面向空间NPU的端到端LLM部署，提出Agent技能系统实现从人工指导到自主运行，架构与机器学习协同创新。
arXiv:2606.07586v1 Announce Type: new Abstract: Spatial neural processing units (NPUs) provide an energy-efficient platform for edge LLM inference, bu…
20
End-to-End Training for Discrete Token LLM based TTS System
提出端到端训练方法，将离散token LLM与TTS系统结合，提升语音合成效果。
arXiv:2606.09234v1 Announce Type: cross Abstract: Recent state-of-the-art (SOTA) text-to-speech (TTS) systems typically adopt a cascaded pipeline cons…