提出算法催化热力学理论，揭示智能计算的能耗极限与信息代价。
Watts-per-Intelligence Part II: Algorithmic Catalysis
- 通过信息热力学分析，发现可复用计算结构能减少不可逆操作。
- 证明加速上限由底物与任务描述间算法互信息决定，且编码有最小能耗。
- 适用于理解大模型能耗瓶颈，适合关注高效智能系统的研究者。
我们在瓦特每智能框架内发展了算法催化热力学理论，识别出一类任务中可重复使用的计算结构，能在满足有限恢复和结构选择性约束的前提下减少不可逆操作。我们证明，任何特定任务类的加速上限由底物与类描述符之间的算法互信息决定，而编码该信息需付出最低热力学成本，即兰道尔擦除代价。结合上述结果，得到一个耦合定理，可下界估计算法催化剂在能量上具有优势所需的部署周期。该框架以仿射SAT类为例进行演示，并将现有学习系统置于智能计算的信息热力学约束之下。
原文摘要 · Abstract (English)
We develop a thermodynamic theory of algorithmic catalysis within the watts per intelligence framework, identifying reusable computational structures that reduce irreversible operations for a task class while satisfying bounded restoration and structural selectivity constraints. We prove that any class specific speed-up is upper-bounded by the algorithmic mutual information between the substrate and the class descriptor, and that encoding this information incurs a minimum thermodynamic cost via Landauer erasure. Combining these results yields a coupling theorem that lower-bounds the deployment horizon required for an algorithmic catalyst to be energetically favourable. The framework is illustrated on an affine SAT class and situates contemporary learned systems within an information thermodynamic constraint on intelligent computation.
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。