Skip to content
Discussion options

You must be logged in to vote

这个没救的,很多模型都有同样的毛病。网上方法有很多,本地模型可以调整温度、重复惩罚那些。调api的话,除了系统提示词做一些改进外,最好的手段是,一旦发生let me do this,找到源头,分叉这个对话。
原理比较蛋疼,就是大模型的输出是概率,一旦被这些废话污染了上下文,ai在推理时,会变得优先选择说废话。

Replies: 3 comments 2 replies

Comment options

You must be logged in to vote
0 replies
Comment options

You must be logged in to vote
0 replies
Comment options

You must be logged in to vote
2 replies
@LiJoeAllen
Comment options

@oldzhang82
Comment options

Answer selected by LiJoeAllen
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
2 participants