AI Alignment Forum
reasoning
to
opaque
models
of
我们如何在推理不透明的模型上进行 SFT？
However, reasoning in interpretable language might turn out to be uncompetitiveif so, it seems probable that opaque reasoning will be adopted in frontier AI labs。Its not obvious that well be able to do training that affects model reasoning like this if models have opaque reasoning though, because we...
2026-02-21
阅读全文
