Neuro-symbolic framework that uses LLM to generate G-code from CAD models and separation logic to verify and iteratively correct spatial collisions via bounding-box feedback.
Habilitationsschrift, Technical University of Kaiserslautern, Kaiserslautern, Germany
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Correct-by-Construction G-Code Generation: A Neuro-Symbolic Approach via Separation Logic
Neuro-symbolic framework that uses LLM to generate G-code from CAD models and separation logic to verify and iteratively correct spatial collisions via bounding-box feedback.