Skip to content

关于“轨迹探索”的一点疑问 #23

Description

@rebeccawn

文章中3.2章节提到“Subsequently, these synthetic instructions T are executed in environment E by a model like GPT-4o, producing a complete set of trajectories, denoted as G = {g1, g2, . . . , gN }.”
请问生成的轨迹中包含的三元组和低级指令,是新产生的,还是和逆向合成阶段的信息相关联的。如果是新产生的话,是不是可以理解为逆向合成只达到一个生成高级指令的目的?
期待您的回复!

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions