Online local training improves generative thermodynamic computers efficiency
Online local learning for generative thermodynamic computing
Machine Learning
Summary
Thermodynamic computers use heat noise to generate structured data, but training them can be complex. The authors show a new way to train these systems step-by-step instead of waiting until the whole process finishes, making training more efficient. They tested this on handwritten digit data and found their method uses less heat overall, even with some noise in updates. Their work also explains how different types of noise and precision affect these systems.
What this means in practice
- •For machine learning engineers: Train generative thermodynamic computing models more efficiently using stepwise local updates that reduce heat dissipation and maintain data quality on standard image data.
- •For hardware accelerator designers: Design thermodynamic computer hardware with lower precision storage for trained parameters while maintaining performance, optimizing for reduced energy consumption during inference.
Authors
Huilin Wang, Weibing Deng
Abstract
Generative thermodynamic computers turn thermal noise into structured data through Langevin dynamics. We train these systems with a local update at each integration step. The reverse-path Onsager-Machlup objective yields a coupling gradient that is a symmetric sum of local residual-state correlations. We apply this gradient immediately rather than accumulating it over a full trajectory. In digital simulations using MNIST prototypes, online and trajectory-batch training reach similar validation losses on fixed noising paths. Models trained online release less heat on average in all five independently seeded pairs, with both models' parameters held fixed during sampling. Auxiliary classifier and nearest-prototype measures change modestly, while pairwise diversity decreases. The response to noise depends strongly on where the errors enter: independent zero-mean errors in the formed updates produce little heat change over a finite range of noise amplitudes, whereas residual offset and temporal correlation have much larger effects. Storing trained couplings requires substantially less precision than resolving deterministic updates during training. Together, these results establish a local online training method and show how update timing, noise structure, and precision affect generative thermodynamic computing.