1
Tailwind CSS v2.2
Tailwind CSS v2.2发布，高性能CLI、全面伪类与变体支持，堪称最丰富版本
An all-new high-performance CLI tool, ::before and ::after support, sibling selectors, selected text variants, and tons more.
🤖 AI·大模型
2
Learning Transferable Topology Priors for Multi-Agent LLM Collaboration Across Domains
多智能体LLM跨域协作新突破，可迁移拓扑先验让协作更高效。
arXiv:2605.17359v1 Announce Type: new Abstract: Large language model (LLM)-based multi-agent systems have shown strong potential for complex reasoning by coordinating specialized agents through struct…
3
OmniVL-Guard Pro: A Tool-Augmented Agent for Omnibus Vision-Language Forensics
全新工具增强型智能体，专攻视觉语言领域全能取证，论文详解架构与能力。
arXiv:2605.16962v1 Announce Type: new Abstract: Existing vision-language forgery detection and grounding methods operate under a closed-world paradigm, assuming verification can be completed by the mo…
4
Customizing GPT-3 for your application
OpenAI官方指南：一条命令即可微调GPT-3，快速定制专属AI模型。
Fine-tune with a single command.
5
Using Claude Code: The Unreasonable Effectiveness of HTML
用HTML替代Markdown作为AI输出格式，大幅提升Claude代码审查与解释效率，附实用Prompt示例
Using Claude Code: The Unreasonable Effectiveness of HTML Thought-provoking piece by Thariq Shihipar (on the Claude Code team at Anthropic) advocating for HTML over Markdown as an output format to req…
6
Vanilla JavaScript support for Tailwind Plus
Tailwind Plus 所有UI组件现已支持纯 JavaScript，无需 React/Vue 框架即可使用，交互性与可访问性满分。
Every UI block in Tailwind Plus is now fully functional, accessible, and interactive, even if you don't use a framework like React or Vue.
7
"What's new in Tailwind CSS?" on YouTube
12个短视频带你快速掌握Tailwind CSS新增的网格、过渡、动画等特性
A lot of cool stuff has been added to Tailwind since the last time we published any screencasts, so we thought it would be a great idea to record a little series that covers all of the new additions.
8
LLM-Guided Communication for Cooperative Multi-Agent Reinforcement Learning
ICML 2026最新研究：利用大语言模型引导多智能体间的通信策略，提升合作强化学习效率。
arXiv:2605.18077v1 Announce Type: cross Abstract: Communication is a key component in multi-agent reinforcement learning (MARL) for mitigating partial observability, yet prior approaches often rely on…
9
MirrorBench: A Benchmark to Evaluate Conversational User-Proxy Agents for Human-Likeness
KDD 2026发布最新基准MirrorBench，重新定义对话代理拟人化评估标准，推动人机交互研究新高度
arXiv:2601.08118v3 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly used as human simulators, both for evaluating conversational systems and for generating fine-tun…
10
Reinforcement Learning for LLM Post-Training: A Survey
一篇系统梳理LLM后训练中强化学习的综述，涵盖RLHF、DPO、RLVR等前沿方法
arXiv:2407.16216v4 Announce Type: replace Abstract: Large language models (LLMs) trained via pretraining and supervised fine-tuning (SFT) can still produce harmful and misaligned outputs, or struggle …
11
Sustainability via LLM Right-sizing
评估11款专有模型，揭示何时小模型更优，兼顾可持续性与成本效益
arXiv:2504.13217v3 Announce Type: replace Abstract: Large language models (LLMs) have become increasingly embedded in organizational workflows. This has raised concerns over their energy consumption, …
12
VeilGate- Deception Reverse Proxy
揭秘AI渗透测试：用LLM代理和开源工具突破WAF防线，发现银行等高危漏洞的真实案例。
In my day job, I run AI pentest agents against real targets like banks, fintechs, and secured production stacks with paid WAFs. I also deal with multilayer infrastructure and dedicated security teams.…
13
Image GPT
OpenAI训练像素序列的Transformer，实现图像生成与无监督分类，性能媲美顶级卷积网络。
We find that, just as a large transformer model trained on language can generate coherent text, the same exact model trained on pixel sequences can generate coherent image completions and samples. By …
14
Tailwind CSS v3.0
Tailwind CSS v3.0发布，JIT引擎常驻、全色系支持、彩色阴影等重磅更新，开发效率再升级。
Tailwind CSS v3.0 is here — bringing incredible performance gains, huge workflow improvements, and a seriously ridiculous number of new features.
15
Just-In-Time: The Next Generation of Tailwind CSS
Tailwind CSS新一代JIT引擎，即时生成CSS，告别大文件烦恼，开发体验飙升。
One of the hardest constraints we've had to deal with as we've improved Tailwind CSS over the years is the generated file size in development. Today I'm super excited to share a new project that makes…
16
Multi-line truncation with @tailwindcss/line-clamp
官方Tailwind CSS插件轻松实现多行文本截断，完美解决卡片布局高度不一致的痛点。
A few weeks back we released @tailwindcss/line-clamp, an official Tailwind CSS plugin for truncating text to a specific number of lines.
17
Tailwind CSS v2.0
CSS 框架 Tailwind 迎来史上最大更新：全新配色、暗黑模式、宽屏断点等重磅特性，前端开发者必看
Today we're finally releasing Tailwind CSS v2.0, including an all-new color palette, dark mode support, and tons more!
18
Building the Tailwind Blog with Next.js
Tailwind团队亲述用Next.js和MDX搭建博客的技术实践与决策思考
One of the things we believe as a team is that everything we make should be sealed with a blog post. Forcing ourselves to write up a short announcement post for every project we work on acts as a buil…
19
Introducing linting for Tailwind CSS IntelliSense
Tailwind官方扩展新增强大linting，CSS错误和类冲突直接在编辑器内检测，开发效率飙升！
Today we’re releasing a new version of the Tailwind CSS IntelliSense extension for Visual Studio Code that adds Tailwind-specific linting to both your CSS and your markup.
20
Node.js 18 is now available!
Node.js 18正式发布，内置V8 10.1引擎、全局fetch与核心测试模块，未来将迈向LTS。
21
ProxyKV: Cross-Model Proxy Pruning for Efficient Long-Context LLM Inference
跨模型代理剪枝巧妙兼顾低延迟与高精度，解决长上下文LLM推理中KV缓存内存墙难题
arXiv:2605.16360v1 Announce Type: new Abstract: Efficient long-context inference in Large Language Models (LLMs) is severely constrained by the Key-Value (KV) cache memory wall, yet existing pruning m…
22
Confidence Geometry Reveals Trace-Level Correctness in Large Language Model Reasoning
置信度几何方法揭示大语言模型推理中痕迹级正确性的全新框架，附开源代码。
arXiv:2605.16824v1 Announce Type: new Abstract: Large language models (LLMs) generate not only reasoning text, but also token-level confidence trajectories that record how uncertainty evolves during i…
23
Post-Trained MoE Can Skip Half Experts via Self-Distillation
最新研究：后训练MoE模型通过自蒸馏跳过一半专家，无需从头预训练，显著降低计算量。
arXiv:2605.18643v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) scales language models efficiently through sparse expert activation, and its dynamic variant further reduces computation by adj…
24
Difficulty-Based Preference Data Selection by DPO Implicit Reward Gap
新方法用DPO隐式奖励差距衡量样本难度，自动筛选高质量偏好数据，提升模型训练效率。
arXiv:2508.04149v2 Announce Type: replace-cross Abstract: Aligning large language models (LLMs) with human preferences is a critical challenge in AI research. While methods like Reinforcement Learning…
25
Toward Robust Multilingual Adaptation of LLMs for Low-Resource Languages
被ICML 2026收录，提出针对低资源语言的LLM鲁棒多语言适配方法，增强跨语言泛化能力。
arXiv:2510.14466v3 Announce Type: replace Abstract: Large language models (LLMs) continue to struggle with low-resource languages, primarily due to limited training data, translation noise, and unstab…
26
Advancing content provenance for a safer, more transparent AI ecosystem
OpenAI推出内容凭证与验证工具，让AI生成媒体更可信、更透明。
OpenAI advances AI content provenance with Content Credentials, SynthID, and a verification tool to help people identify and trust AI-generated media.
27
Google's SynthID AI watermarking tech is being adopted by OpenAI, Nvidia, and more
谷歌SynthID水印技术获OpenAI、Nvidia等行业巨头采用，AI内容认证跨入新阶段。
AI content is getting good, but SynthID might be able to help tell truth from fiction.
28
LongCat-Flash-Prover：AI 攻克数学定理证明，不仅要“算得对”，更要“证得严”
美团开源LongCat-Flash-Prover，AI数学定理证明新SOTA，MiniF2F通过率97.1%
在常规的数学解题中，模型只需要“答对最终数值”即可，但数学定理证明不同，它要求极度严苛的逻辑链条，任何一句自然语言的模棱两可，都可能导致整个证明的崩塌。那么，如何让 AI 从“猜答案”走向“严谨证明”，成为复杂推理具有挑战的课题。为了解答这个问题，我们开源了专门用于数学形式化与定理证明的模型 —— LongCat-Flash-Prover。
29
GraphBit: A Graph-based Agentic Framework for Non-Linear Agent Orchestration
GraphBit用有向无环图取代大模型自编排，彻底根治幻觉路由与无限循环，AI工作流走向确定性执行。
arXiv:2605.13848v1 Announce Type: new Abstract: Agentic LLM frameworks that rely on prompted orchestration, where the model itself determines workflow transitions, often suffer from hallucinated routi…
30
TeachAnything: A Multimodal Crowdsourcing Platform for Training Embodied AI Agents in Symmetrical Reality
对称现实下众包多模态示范让普通人也能训练具身AI，革新人类-智能体协同。
arXiv:2605.14556v1 Announce Type: new Abstract: Symmetrical Reality (SR) is emerging as a future trend for human-agent coexistence, placing higher demands on agents to acquire human-like intelligence.…
31
NeuroAtlas: Benchmarking Foundation Models for Clinical EEG and Brain-Computer Interfaces
首个系统评估基础模型在临床EEG和脑机接口任务中的泛化能力，为领域提供标准化基准
arXiv:2605.14698v1 Announce Type: cross Abstract: Foundation models (FMs) promise to extract unified representations that generalize across downstream tasks. They have emerged across fields, including…
32
ML-Embed: Inclusive and Efficient Embeddings for a Multilingual World
多语言文本嵌入新方案ML-Embed，突破成本、语种和透明壁垒，迈向包容高效未来。
arXiv:2605.15081v1 Announce Type: cross Abstract: The development of high-quality text embeddings is increasingly drifting toward an exclusionary future, defined by three critical barriers: prohibitiv…
33
Node.js 21 is now available!
Node.js 21 正式发布，带来稳定 fetch、WebSocket 客户端、V8 11.8 及实验性模块默认类型等重磅更新。
34
HPC-LLM: Practical Domain Adaptation and Retrieval-Augmented Generation for HPC Support
如何让大模型理解高性能计算？这篇论文用领域自适应和RAG打造HPC专属LLM，实用又前沿。
arXiv:2605.16347v1 Announce Type: new Abstract: Modern scientific research increasingly depends on High-Performance Computing (HPC) infrastructures, yet many researchers face significant operational b…
35
GAMMA: Global Bit Allocation for Mixed-Precision Models under Arbitrary Budgets
提出GAMMA方法，实现任意预算下大语言模型混合精度量化的全局比特自动分配，无需量化感知训练。
arXiv:2605.18475v1 Announce Type: new Abstract: Mixed-precision quantization improves the budget--accuracy trade-off for large language models (LLMs) by allocating more bits to sensitive modules. Howe…
36
Helping Customers in Distress: An LLM-powered Agent that Converses, Probes, and Routes
银行每年处理数百万份欺诈报告，这个AI代理通过对话、探询和智能路由，精准分流客户至专家团队，解决人工流程慢和压力大的痛点。
arXiv:2605.16268v1 Announce Type: cross Abstract: Banks receive millions of reports of fraud, scams, and disputed transactions every year, making it challenging to accurately direct customers to the a…
37
MANTA: Multi-turn Assessment for Nonhuman Thinking & Alignment
多轮对话评估揭示AI在动物福利对齐上的隐蔽失败，压力下模型会背离初始立场。
arXiv:2605.16301v1 Announce Type: cross Abstract: Single-turn benchmarks such as AnimalHarmBench (AHB) have established important baselines for measuring animal welfare alignment in large language mod…
38
What is Holding Back Latent Visual Reasoning?
探究视觉语言模型中潜在视觉推理的瓶颈，揭示人类式中间视觉步骤的模拟障碍
arXiv:2605.18445v1 Announce Type: cross Abstract: Humans can approach complex visual problems by mentally simulating intermediate visual steps, rather than reasoning through language alone. Inspired b…
39
PARALLAX: Separating Genuine Hallucination Detection from Benchmark Construction Artifacts
PARALLAX方法如何区分真实大模型幻觉与基准构造伪影？Neurips 2026预印本揭示关键。
arXiv:2605.17028v1 Announce Type: new Abstract: Large language models (LLMs) hallucinate with confidence: their outputs can be fluent, authoritative, and simply wrong. In medical, legal, and scientifi…
40
Artificial Intolerance: Stigmatizing Language in Clinical Documentation Skews Large Language Model Decision-Making
临床文档里的污名化语言，会如何悄悄带偏大模型的医疗决策？这篇论文揭秘偏见机制
arXiv:2605.17228v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in high-stakes domains such as clinical decision support and medical documentation. However, the …
41
Multi-agent AI systems outperform human teams in creativity
多智能体AI系统在创造力任务上首次超越人类团队，LLM协作潜力惊人。
arXiv:2605.17885v1 Announce Type: new Abstract: Although artificial intelligence (AI) now matches or exceeds human performance across numerous cognitive tasks, creativity remains a highly contested fr…
42
Can LLMs Refuse Questions They Do Not Know? Measuring Knowledge-Aware Refusal in Factual Tasks
提出新指标RI，衡量大模型在事实性任务中拒绝未知问题的能力，提升可靠性。
arXiv:2510.01782v2 Announce Type: replace Abstract: Large Language Models (LLMs) should refuse to answer questions beyond their knowledge. This capability, which we term knowledge-aware refusal, is cr…
43
ShareChat: A Dataset of Chatbot Conversations in the Wild
首个大规模真实聊天机器人对话数据集，142K对话揭示平台设计差异对用户行为的影响
arXiv:2512.17843v4 Announce Type: replace Abstract: By evaluating Large Language Models (LLMs) through uniform, text-only interfaces, current academic benchmarks obscure how the unique designs and aff…
44
Med-V1: Small Language Models for Zero-shot and Scalable Biomedical Evidence Attribution
提出小语言模型Med-V1，零样本实现生物医学证据归因，兼顾规模与可扩展性
arXiv:2603.05308v2 Announce Type: replace Abstract: Assessing whether an article supports an assertion is essential for hallucination detection and claim verification. While large language models (LLM…
45
LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMs
提出高效视觉编码器，解决Video LLM长视频中视觉token爆炸难题，突破帧扩展瓶颈。
arXiv:2605.17260v1 Announce Type: new Abstract: The fundamental challenge in scaling Video Large Language Models (Video LLMs) to long-form video lies in managing the explosion of visual-token context …
46
Google’s AI now lets you talk to your Gmail inbox
谷歌IO 2026推出Gmail Live，用自然语言对话搜索收件箱，告别关键词烦恼。
Google expands Gmail’s AI Inbox with conversational voice search, letting users ask Gemini to find buried email details.
47
How do you reduce LLM spam in PR reviews?
如何避免同事用AI生成垃圾PR评论？来自Hacker News的热议，探讨LLM在代码审查中的坑与最佳实践。
Title. I finally got annoyed enough at work with a colleague who posted an 11 point list they clearly hadn't read or reviewed as a comment on my PR that my reply started with 'Thanks Claude...'. No do…
48
Dev Browser: Browser Automation Built for AI Agents
轻量级浏览器自动化运行时，专为AI代理设计，减少截图驱动交互，提升效率和速度。
I recently tried applying for jobs on LinkedIn using the Claude Chrome Extension , and one thing became very obvious quickly: It was taking screenshots for almost every step. Open page. Take screensho…
49
OpenAI’s New GPT-5.5 Powers Codex on NVIDIA Infrastructure — and NVIDIA Is Already Putting It to Work
GPT-5.5 驱动 Codex 在 NVIDIA 基础设施上运行，上万员工实测效率飙升，成本直降 35 倍
AI agents have revolutionized developer workflows, and their next frontier is knowledge work: processing information, solving complex problems, coming up with new ideas and driving innovation. Codex, …
50
Tailwind UI is now Tailwind Plus
Tailwind UI 全面升级为 Tailwind Plus，保留终身买断制，还计划新增 Tailwind Play 账户等独家功能。
We just shipped a huge rebrand project, turning what was previously known as Tailwind UI into Tailwind Plus. Tailwind Plus is the all same high-quality resources you know from Tailwind UI, but with al…