DuoPlus雲手機

別錯過 #DuoPlus 萬聖節限定活動！ 兩大福利等你來拿！ 🎃 輸入萬聖專屬折扣碼 Helloween15 享85折優惠 🎁 抽獎活動開啟！關注 @DuoPlusofficial，留言或轉發就能參加；TG社群另有獨立抽獎，透過X主頁連結進群就能參與 獎池內容： 🥇 $20 雲手機贈送金 + 1 組美國雲號碼 🥈 6 個月香港／印尼雲號碼 🥉 $10雲手機贈送金 💡備註：贈送金僅可用於購買雲手機與開機時長 ⏰ 抽獎截止：10 月 30 日 17:00 (UTC+8) 👉 立即領取折扣：duoplus.cn/share/tuite2

While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected. Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions. Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent. But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models? We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem. Read the full report: cohere.com/blog/concerning-b…

We have continued to dig into the root cause here and would like to offer an update. Our engineering team discovered that the spike that triggered the incorrect enforcement was primarily in the account’s “reading” behavior. Specifically, our network security systems detected the use of a Claude tool reading profiles, posts, and hashtags, the rate of which sharply increased (~4x volume) in intense bursts on October 17. These abrupt, high-volume, sequential requests in a short period resembled automated bot traffic rather than an ordinary human browsing our systems. This activity caused our automated protections to incorrectly take the No Kings Instagram account offline. An automated re-review of the removal action an hour later recognized that shouldn’t have happened and restored the account. It unfortunately took additional time for us to straighten out the situation with admin access to the IG account. We want to reiterate our sincere apology for the error.

Welcome HomeView

Introducing dots, powered by GPT-6 Astra. Remarkably capable, always-on agents built to handle everything.

Grok 4.7 is here. It's a notable improvement over Grok 4.6 at the same price and speed.