用极小网络递归推理，轻松超越大模型解难题。
Less is More: Recursive Reasoning with Tiny Networks
- 单个两层小网络循环推理，机制简单却高效。
- 仅700万参数，在ARC-AGI任务上达45%准确率。
- 适合资源受限场景，对小模型研究者极具启发。
层级推理模型（HRM）采用两个小型神经网络以不同频率递归运行，受生物启发，能在仅2700万参数和约1000个训练样本下，超越大型语言模型（LLMs）在数独、迷宫和ARC-AGI等难题上的表现。然而其原理尚不清晰且可能非最优。本文提出更简单的极小递归模型（TRM），仅用一个两层的小网络实现显著更强的泛化能力。TRM仅700万参数，在ARC-AGI-1上取得45%测试准确率，在ARC-AGI-2上达8%，远超多数大模型（如Deepseek R1、o3-mini、Gemini 2.5 Pro），且参数量不足其0.01%。
原文摘要 · Abstract (English)
Hierarchical Reasoning Model (HRM) is a novel approach using two small neural networks recursing at different frequencies. This biologically inspired method beats Large Language models (LLMs) on hard puzzle tasks such as Sudoku, Maze, and ARC-AGI while trained with small models (27M parameters) on small data (around 1000 examples). HRM holds great promise for solving hard problems with small networks, but it is not yet well understood and may be suboptimal. We propose Tiny Recursive Model (TRM), a much simpler recursive reasoning approach that achieves significantly higher generalization than HRM, while using a single tiny network with only 2 layers. With only 7M parameters, TRM obtains 45% test-accuracy on ARC-AGI-1 and 8% on ARC-AGI-2, higher than most LLMs (e.g., Deepseek R1, o3-mini, Gemini 2.5 Pro) with less than 0.01% of the parameters.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。