眼镜巨头EssilorLuxottica收购法国XR初创Lynx，其未上市的混合现实头显Lynx-R2命运成谜，背后或与Meta合作有关。
IT之家 7 月 16 日消息，法国 XR 硬件初创公司 Lynx 创始人兼 CEO Stan Larroque 向支持者发送公开邮件，确认公司已正式被全球眼镜巨头 EssilorLuxottica（依视路陆逊梯卡）收购。 Larroque 在邮件中表示，EssilorLuxottica 将获得 L…
用Cloudflare AI Search和R2搭建可搜索的代码库或文档库，支持/ask和/mcp端点，适合开发者快速集成
Hello, So I got sick of my agents not having ready access to information they needed about projects that I'm working on ready and at hand so I built a…
Rivian上调全年电动车销售预期，Q2交付量超预期，EDV和R1增长强劲，R2开始交付。
The company now expects to ship a few thousand more vehicles by the end of 2026 than it previously expected, after launching its R2 SUV last month.
针对MLLMs视觉STEM感知难题，CodePercept创新引入代码接地机制，为多模态理解开辟新路径，已被CVPR2026收录。
arXiv:2603.10757v2 Announce Type: replace Abstract: When MLLMs fail at Science, Technology, Engineering, and Mathematics (STEM) visual reasoning, a fu…
ICLR2026最佳论文，用结构不确定性量化LLM逻辑推理一致性，为评估推理可靠性开辟新视角。
arXiv:2606.17312v1 Announce Type: new Abstract: Large language models can arrive at the same answer through reasoning paths that are unstable, contrad…
Rivian刚交付热门SUV R2就宣布裁员数百人，累计亏损300亿，2027盈利目标承压。
The company said the cuts were part of a restructuring meant to help scale to profitability. Rivian recently pushed back its profitability goal to inv…
用两个词就能表达任意分割掩码，SAM最新变体实现高效语义token化表示。
arXiv:2601.16093v2 Announce Type: replace Abstract: Pixel-wise capabilities are essential for building interactive intelligent systems. However, pixel…
丰田 MR2 原型车曝光：轻量化中置后驱，2028 年上市，驾驶乐趣超 GR Yaris。
IT之家 6 月 12 日消息，汽车媒体 autoexpress 昨日（6 月 11 日）发布博文，试驾了丰田新款轻量化中置跑车，该车型预估 2028 年上市， 上市后可能名为 MR2 或 Celica。 IT之家援引博文介绍，该轻量化跑车项目由公司董事长丰田章男亲自领导，开发思路深受赛车运动影响，…
Rivian R2首次试驾体验：烧掉250亿美元后的救命稻草，重新定义电动SUV。
With a competitive price, winning design, and better performance than the R1, Rivian could be set to break into the big leagues. Just make sure you ge…
Rivian R2加长轴距颠覆传统SUV设计，后座空间与全降车窗带来惊喜，首驾体验揭示纯电新标杆。
Rivian's second EV is the sub-$60,000 R2, and it was worth the wait.
提出R2IF框架，用复合奖励让大模型函数调用时推理与决策对齐，提升可解释性。
arXiv:2604.20316v2 Announce Type: replace Abstract: Function calling empowers large language models (LLMs) to interface with external tools, yet exist…
用预测性路由重放技术优化MoE大模型强化学习，显著提升训练效率与模型性能
arXiv:2606.00395v1 Announce Type: new Abstract: Mixture of Experts (MoE) Large Language Models (LLMs) achieve strong performance at scale. However, re…
探讨仅用最后一层进行不确定性量化是否足够，挑战主流认知，对模型可靠部署有重要启发
arXiv:2605.30741v1 Announce Type: cross Abstract: Epistemic uncertainty quantification (UQ) for deep neural networks (DNNs) is a requirement for safe …
CVPR 2026最新研究：通过早期Token压缩实现快速视频理解，颠覆传统视频处理效率瓶颈。
arXiv:2605.30010v1 Announce Type: new Abstract: Video large language models (Video-LLMs) have demonstrated strong capabilities in video understanding …
来自ICLR2026的研究：强化学习让推理更紧凑，监督微调则让推理更发散，揭示两种范式在大模型推理上的本质差异。
arXiv:2509.21128v2 Announce Type: replace Abstract: Large language models (LLMs) are typically trained by reinforcement learning (RL) with verifiable …
CVPR 2026论文，提出多帧空间理解新方法，刷新多模态大模型空间推理能力。
arXiv:2505.17015v2 Announce Type: replace-cross Abstract: Multi-modal large language models (MLLMs) have rapidly advanced in visual tasks, yet their s…
基于智能体的自动证明优化框架，提升数学证明的简洁性与可读性，ICLR 2025 顶会论文。
arXiv:2410.04753v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have been used to generate formal proofs of mathematical theore…
用反馈控制器实现精准激活引导，为提升大模型可控性提供新思路，ICLR2026论文。
arXiv:2510.04309v3 Announce Type: replace Abstract: Controlling the behaviors of large language models (LLM) is fundamental to their safety alignment …
教你如何让小型语言模型学会判断何时该“求救”，避免盲目依赖昂贵大模型，提升Agent系统效率的突破性研究。
arXiv:2605.16604v1 Announce Type: new Abstract: Efficient agentic systems should incur expensive frontier-model costs only on decisions where a cheape…
8B参数模型专用生物实验protocol生成，步骤顺序不乱、剂量无幻觉，超越GPT-4o，已登ICLR2026。
超越GPT-4o