SurfaceBench: Benchmark for Symbolic Regression on 3D Surfaces

SurfaceBench is a collection of symbolic regression problems focused on 3D surface modeling, created by ishobhnik and last updated in March 2026. The dataset includes 15 different categories of surface types, each with multiple instances. It is structured in HDF5 format and serves as a diverse benchmark for algorithm testing.

Use Cases

• Benchmarking symbolic regression algorithms based on the 15 categories of surface types.
• Developing models for 3D surface function approximation based on the described symbolic regression problems.
• Testing algorithm generalization across diverse geometric instances as mentioned in the dataset structure.

Strengths

• Contains 15 distinct categories of surface types, providing categorical diversity.
• Each category includes multiple instances, offering intra-category variation.
• Dataset is structured in HDF5 format, which is efficient for storing hierarchical data.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats beyond HDF5, and sample data are unknown, which may limit suitability assessment.
• License information is not provided in the available metadata.

Provenance

Source

ishobhnik via Hugging Face.

Collection Method

Likely compiled or generated for benchmarking purposes.

Time Range

null

Freshness

Last updated 2026-03-23 05:22:41; freshness should be verified.

Geography

null