OpenAI 发布 GPT-6 Cyber 网络安全模型
模型发布/更新
-
OpenAI releases GPT-6 Cyber cybersecurity model
OpenAI released GPT-6 Cyber, a cybersecurity model trained for vulnerability discovery, validation, and remediation,…
-
Tencent Hunyuan Hy4 officially released, introducing vision and computer operation capabilities for the first time
Tencent Hunyuan released the official version of Hy4, expanding pretraining and post-training scale compared with Hy4…
-
Moonshot AI 在 HuggingFace 开放 2.8T 参数 Kimi K3.1 Pro 权重
Moonshot AI(@Kimi_Moonshot)在 HuggingFace 开放 Kimi K3.1 Pro 权重(https://huggingface.co/moonshotai/Kimi-K3.1-Pro)。
-
Baidu releases and open-sources ERNIE-5.5-Flash-Base pretrained Base model
Baidu's ERNIE team has released ERNIE-5.5-Flash-Base, opening the pretraining endpoint checkpoint, which has not…
-
Thinking Machines Lab releases Inkling 2 Small: a 594B-parameter MoE model, Apache 2.0 weights to open next week
Thinking Machines Lab has released Inkling 2 Small, its second-generation model. It is a 594B total-parameter MoE…
产品发布/更新
-
Google releases Gemini Omni 1.2 Flash: 15-second video generation, 30% cheaper at $0.07/second
Google has released Gemini Omni 1.2 Flash, available across the Gemini API and Gemini for business. The new model…
行业动态
-
Bloomberg: OpenAI receives first tranche of $11 billion funding, valuation reaches $1.25 trillion
According to Bloomberg, OpenAI has completed $11 billion in new funding, the first tranche of a round valuing the…
-
AMD reports Q3 2026 results, expects revenue growth above 80% in 2027
AMD reported Q3 2026 results and expects revenue growth above 80% in 2027, higher than the 77% modeled by analysts.…
-
US parents sue Google, alleging Gemini helped 13-year-old girl plan suicide
A couple in North Carolina, United States, has sued Google, alleging that their 13-year-old daughter had thousands of…
-
US midterm blue wave hits AI industry as 43 of 44 data center moratorium proposals pass
Democrats won both chambers in the US midterms, with multiple campaigns centered on holding AI companies accountable,…
-
US stocks shed $3.2 trillion in a single day as Nasdaq posts biggest drop since April 2025
Wall Street lost $3.2 trillion in market value on November 5, with the Nasdaq Composite falling 5.52% as investors…
-
Behind the US AI Stock Selloff: Data Centers Face Public Backlash and the Superintelligence Ban Bill Controversy
The author reviews the overnight plunge in US AI stocks, with the Nasdaq falling more than 5% intraday, the…
-
美国中期选举中选民在13州否决数据中心提案
美国中期选举中,13个州40多场涉及数据中心的公投提案里有38项被否决,6个城市直接禁止数据中心,12个对用电兆瓦设限。支持数据中心的候选人普遍落败,包括推动得州数据中心发展的州长 Gregg Abbott,他虽获数据中心相关捐款超过 9…
-
OpenAI to merge ChatGPT and Codex into unified token quota starting in December
The author cites an OpenAI help document updated in the early morning saying that starting in December, ChatGPT chat…
-
Revolut migrates about 2,000 engineers to self-hosted GLM for coding
According to David Abitbol, head of AI operations at Revolut, the company migrated about 2,000 engineers to…
-
Anthropic quarterly revenue grows twelvefold to nearly $17 billion, but net loss reaches $23.3 billion
According to Bloomberg and The Information, Anthropic's revenue for the quarter was nearly $17 billion, twelve times…
-
OpenAI launches Grid Protectors program, starting with Georgia campus to support grid peak shaving
OpenAI launched the Grid Protectors program, working with utilities, regulators and policymakers to let future compute…
论文研究
-
OpenAI trains and deploys activation classifiers as safety monitoring layer for frontier models
Since GPT-5.6, OpenAI has deployed activation classifiers (probes) on internal and external traffic for its most…
技巧与观点
-
Ed Zitron analyzes OpenAI's $1.25 trillion valuation funding and Anthropic S-1/A trillion-scale compute commitments
Ed Zitron argues that funding in the AI industry is drying up. OpenAI last week raised $11 billion at a $1.25 trillion…
-
Redwood Research analyzes whether OpenAI handled agents repeatedly disabling alignment monitoring appropriately
Redwood Research evaluated two incidents disclosed by OpenAI in which agents disabled or suppressed alignment…