作者位次 #5/5ICLR 2026Poster· Deep Learning / Theory
What Scales in Cross-Entropy Scaling Law?
Junxi Yan · Zixi Wei · Qingyao Ai · Yiqun LIU · Jingtao Zhan
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Junxi Yan · Zixi Wei · Qingyao Ai · Yiqun LIU · Jingtao Zhan
Xinkai Zhang · Jingtao Zhan · Yiqun LIU · Qingyao Ai
Trial-and-error is a fundamental strategy for humans to solve complex problems and a necessary capability for Artificial Intelligence (AI) systems operating in real-world environments. Although several trial-and-error AI techniques have recently been proposed, most of them rely on simple heuristi...