A framework extracts physics priors via LLMs, distills them through a Graph-Masked Attention teacher into a fast student model, and shows high accuracy plus fault tolerance across five manufacturing processes.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Physics-Distilled Neural Network enabled by Large Language Models for Manufacturing Process-Property Predictive Modeling
A framework extracts physics priors via LLMs, distills them through a Graph-Masked Attention teacher into a fast student model, and shows high accuracy plus fault tolerance across five manufacturing processes.