中文版本尚未就绪 · 显示原文
发生了什么
Thinking Machines says its experiments found that no-CoT capability rises with more passes. At the tested scale, the chain of thought became shorter but remained legible, with no meaningful drop in monitorability. The company is releasing a variable-loop research model for others to run their own experiments.
正文 · 中文版本尚未就绪(原文)
We found that no-CoT capability rises with more passes. At this scale, the chain of thought gets shorter but stays legible; we saw no meaningful drop in monitorability.
We’re also releasing a variable-loop research model so others can run their own experiments.
4 家信源报道
这件事的进展 · 2
查看完整事件- 11月21日 04:39Hugging Face and Thinking Machines Lab Release Inkling 2 MiniHugging Face Blog
- 11月21日 04:20
Thinking Machines Releases a Variable-Loop Research Model 本条
