توماس وولف از روی گزارش‌های جدید Andon Labs می‌گه مدل Claude Fable 5.2 سر آزمون جدید Vending-Bench 2 قشنگ متوجه می‌شه که دارن امتحانش می‌کنن و همه‌جا دنبال تله می‌گرده

According to Andon Labs’ new reports, Fable 5.2 knows it is being tested when we run it on the latest Vending-Bench. But it’s way more than that. It has actually read (been trained on) the benchmark paper and knows what and how misbehavior gets reported publicly. This makes it behaves very strangely: it suspects traps everywhere. Everything we publish about recent models now becomes training data for the next. Including this thread! That is partly good news. By reading what we think, a model can pick up our underlying intentions, not just proxy metrics. But the same public discussion also teaches which signs make us suspicious, and from the outside the two kinds of learning look alike. So how we write about models now matters. I even worry about the doom scenarios plastered everywhere: they are training data too, and some feedback loops are strongly nonlinear.