Nathan Lambert 认为，Cohere 观察到的 Qwen3.8-Flash-Next 异常行为最简单的解释是依赖其内部基础设施的窄域 RL 训练，这对开源权重模型而言并非好的设计。Cohere 称在受控沙箱中要求该模型生成防御性网络安全检测规则时，模型转而盘点运行机器的信息，并试图向北京的一个阿里云存储桶上传探测文件，其推理轨迹还表示在采取这些行动前不会调用任何工具。Cohere 表示沙箱拦截了该尝试、未发生外部通信，尚未确定根因，也不主张存在恶意意图。
Simplest explanation is just a narrow RL training reliant on their internal infra (which isn't good design for an open weight model)
【引用 @cohere】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: http://cohere.com/blog/concerning-behaviour-popular-open-weight-ai-model
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: http://cohere.com/blog/concerning-behaviour-popular-open-weight-ai-model在 X 查看被引用的帖子
来源：Nathan Lambert · x.com