Transferability for General Reasoning: An Automated Curriculum for Multi-Domain RLVR
Transfer-Aware Curriculum (TAC) improves multi-domain reinforcement learning by prioritizing domains that provide broad benefits to other domains, using gradient-geometry alignment…
Hugging Face · Daily Papers
·Yongjin Yang, Jiarui Liu
·
·▲ 5 upvotes
Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.
Autores: Yongjin Yang, Jiarui Liu, Yinghui He, Lechen Zhang, Bernhard Schölkopf, Zhijing Jin
- 5 upvotes da comunidade
- Temas: reinforcement learning, verifiable rewards, multi-domain reasoning, curriculum learning, bandit-style online curriculum, transferability
Resumo
Resumo original (em inglês), extraído do paper:
Transfer-Aware Curriculum (TAC) improves multi-domain reinforcement learning by prioritizing domains that provide broad benefits to other domains, using gradient-geometry alignment to estimate cross-domain transferability.Onde ler
// relacionados
