LEMMA:用学习的符号抽象引导高级数学推理
LEMMA: Bootstrapping High-Level Mathematical Reasoning with Learned Symbolic Abstractions
摘要:
人类通过发展抽象层次来驯服数学推理的复杂性。有了适当的抽象,解决困难的问题,可以表达简洁,从而使他们更容易被发现。在本文中,我们提出了学习数学抽象(LEMMA):一个算法,实现这一想法的强化学习代理在数学领域。LEMMA通过一个抽象步骤增强了专家迭代,在这个步骤中,到目前为止找到的解决方案将根据新的更高级别的操作进行重新访问和重写,然后这些操作将可用于解决新问题。我们评估LEMMA的两个数学推理任务-方程求解和分数简化-在一步一步的方式。在这两个领域中,LEMMA提高了现有代理的能力,解决了更多的问题,并且比训练期间看到的更有效地推广到更难的问题。
英文摘要:
Humans tame the complexity of mathematical reasoning by developing hierarchies of abstractions. With proper abstractions, solutions to hard problems can be expressed concisely, thus making them more likely to be found. In this paper, we propose Learning Mathematical Abstractions (LEMMA): an algorithm that implements this idea for reinforcement learning agents in mathematical domains. LEMMA augments Expert Iteration with an abstraction step, where solutions found so far are revisited and rewritten in terms of new higher-level actions, which then become available to solve new problems. We evaluate LEMMA on two mathematical reasoning tasks--equation solving and fraction simplification--in a step-by-step fashion. In these two domains, LEMMA improves the ability of an existing agent, both solving more problems and generalizing more effectively to harder problems than those seen during training.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。