3 ms·
I recall recent work [ACE](https://www.arxiv.org/abs/2510.04618 https://www.arxiv.org/abs/2510.04618) and [GEPA](https://arxiv.org/abs/2507.19457 https://arxiv.
by Weaver_zhu 1y ago
I recall recent work [ACE](https://www.arxiv.org/abs/2510.04618 https://www.arxiv.org/abs/2510.04618) and [GEPA](https://arxiv.org/abs/2507.19457 https://arxiv.org/abs/2507.19457) where models get improved by adapting and adopting different kinds of prompt. The improvements will be expected to be more generalized than fine-tuning.