A CNN trained contrastively on solutions versus perturbed non-solutions guides LLM synthesis of MiniZinc streamliners, delivering up to 932x geometric-mean speedups on hardened benchmarks.
Pawan Kumar, Emilien Dupont, Francisco J
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.AI 2years
2026 2representative citing papers
citing papers explorer
-
Streamlined Constraint Reasoning via CNN Pattern Recognition on Enumerated Solutions
A CNN trained contrastively on solutions versus perturbed non-solutions guides LLM synthesis of MiniZinc streamliners, delivering up to 932x geometric-mean speedups on hardened benchmarks.
- CVEvolve: Autonomous Algorithm Discovery for Unstructured Scientific Data Processing