推荐m_ke · 2025-11
“Some recent works that people might find interesting:”
把它列为大家可能感兴趣的近期工作
仅 700 万参数的微型递归模型,在 ARC-AGI 上胜过多数大模型
论文针对层级推理模型(HRM)可能并不最优、也尚未被充分理解的问题,提出 Tiny Recursive Model(TRM):只用一个仅 2 层的小网络做更简单的递归推理,泛化能力明显优于 HRM。TRM 仅 7M 参数,在 ARC-AGI-1 上达到 45% 的测试准确率、ARC-AGI-2 上 8%,高于 Deepseek R1、o3-mini、Gemini 2.5 Pro 等多数 LLM,而参数量不到它们的 0.01%。
HN 上的人把它当作 HRM 之后用极小循环网络做推理的代表作,围绕其循环网络结构、与 looped transformer 的关系以及可玩的数独演示展开讨论。
“Some recent works that people might find interesting:”
“You might find it an interesting read.”
“I wanted to get a better understanding of this paper”
“Referring to this paper: Less is More”
“This paper blew up a while ago, and it's a looped transformer architecture (though not like the ones discussed here) that isn't autoregressive”