FIE-500k: A Symbolic Dataset for Large Language Models to Solve Second-Kind Fredholm Integral Equations

The dataset FIE-500k accompanies the paper “A Symbolic Dataset for Large Language Models to Solve Second-Kind Fredholm Integral Equations.” It contains symbolic representations of second-kind Fredholm integral equations. The records are generated using context-free grammars with a carefully defined set of basis functions to ensure coverage of diverse function types. Each entry encodes the essential components of a second-kind Fredholm integral equation. The dataset comprises 500,000 validated records with balanced distributions across function categories. It is intended for applications such as symbolic modeling and solving integral equations with language models.