模型发布/更新
-
Google AI framework AlphaProof Nexus cracks 2 math problems unsolved for 56 years
Google DeepMind launched AlphaProof Nexus, which combines large language models with Lean formal verification and…
-
ModelBest open-sources MiniCPM5-1B: surpasses all sub-2B models on AA-Index and runs on phones and browsers
ModelBest has open-sourced MiniCPM5-1B, its new-generation on-device large language model. With only 1B parameters, it…
-
腾讯混元新翻译模型双榜亮眼,小程序发布
腾讯混元发布翻译模型 Hy-MT2,在 Hugging Face 开源模型趋势排行榜上表现突出:1.8B 版本排名第一,30B-A3B(MoE)版本排名第四,下载量已超 7K。官方同步推出了基于该模型的“腾讯混译”微信小程序,支持语音输入…
产品发布/更新
-
Claude Code launches security vulnerability detection plugin
We released a security guidance plugin for Claude Code that helps identify and fix vulnerabilities while writing code.…
-
Kling AI驱动《大卫之家》实现多项行业首创
Jon Erwin - Wonder Project创始人兼首席创意官,Innovative Dreams首席执行官,亚马逊Prime热门剧集《大卫之家》创作者: “Kling AI是《大卫之家》第一季和第二季背后的核心力量。最让我兴奋的…
行业动态
-
Claude Mythos reportedly solves OpenAI's milestone Erdős problem with a clever, concise proof
Anthropic engineer Sholto Douglas said Claude Mythos solved the Erdős unit distance conjecture problem posed by OpenAI…
-
SynthID watermarking expands partnerships to cover over 100 billion pieces of content
SynthID has watermarked more than 100 billion pieces of content, but transparency requires industry collaboration. We…
-
OpenRouter closes $113M Series B funding
Today we announced the close of a $113M Series B round led by @CapitalGVC. Over the past six months, as AI rapidly…
-
Anthropic appoints KiYoung Choi as Representative Director for Korea
Anthropic announced the appointment of KiYoung Choi as Representative Director for Korea to support its soon-to-open…
-
微软亚洲研究院发起全球AI价值观挑战赛
AI能否理解人类价值观的现实复杂性?微软亚洲研究院的全新全球AI价值观挑战赛,诚邀哲学、伦理、法学和社会科学领域的研究人员共同探索。立即注册:https://msft.it/6012vk8OO
论文研究
-
New KPop method pushes Ring-2.6-1T past 76 on SWE-bench Verified
The team introduced KPop to stabilize agentic reinforcement learning training for large-scale MoE models. It replaces…
技巧与观点
-
Sundar Pichai on AI, the future of search and the changing web
Google and Alphabet CEO Sundar Pichai spoke after Google I/O, reviewing the company's strategic reorganization and…
-
"Father of the lobster" Peter open-sources skill-cleaner to health-check AI agent skills
"Father of the lobster" Peter has open-sourced skill-cleaner, a tool for diagnosing and optimizing AI agent skill…
-
A guide to using Gemini Omni video prompts
Google released a guide to using the video generation capabilities of its multimodal model Gemini Omni. The model can…
-
Project Luxo: crossing the uncanny valley of AI media
Through Project Luxo research, Runway found that AI-generated video has crossed the "uncanny valley." It showed AI…
-
How we isolate and control Claude across products
Anthropic controls the deployment risks of Claude agents through three mechanisms, covering user misuse, anomalous…
-
SenseNova-U1全训练代码开源,支持多模态多任务训练
OpenSenseNova开源了SenseNova-U1的完整训练代码库,支持其8B密集模型与A3B MoE架构。该代码库使用一个统一的框架,可同时训练多种多模态任务,包括文本到图像生成、图像编辑、交错生成及文本与视觉理解。工程上为大规模…
-
Looking ahead: some thoughts for May 2026
The article looks ahead at developments in AI through May 2026. It covers the release of Gemini Flash 3.5, a new…