Agents are increasingly given open-ended research tasks: discovering an empirical law from self-designed experiments, improving a heuristic whose optimum nob…
机构:UMich
来源:arXiv 2610.02588 | AI4Papers 论文推荐平台