symbolic-regression
SymbolicRegressionQA
SymbolicRegressionQA
SymbolicRegressionQA (SRQA) is a multiple-choice visual question answering dataset introduced in
Bayesian Adaptation Gym (UAI 2026). Each question shows a plot of a
symbolic expression and asks which of four candidate formulas describes the relationship between the
variables.
Splits
Split
Examples
train
10,000
validation
1,000
test
1,000
Fields
Field
Type
Description
image
Image
224x224 grayscale… See the full description on the dataset page: https://huggingface.co/datasets/csamplawski/SymbolicRegressionQA.symbolic-regression-surfaces
Comprehensive Symbolic Regression Surface Dataset
Overview
This dataset contains a comprehensive collection of symbolic regression problems focused on 3D surface modeling. The dataset includes 18 different categories of surface types, each with multiple instances, providing a diverse benchmark for symbolic regression algorithms.
Dataset Structure
The dataset is organized in HDF5 format with the following structure:
/
├── Category_1/
│ ├── Instance_1/
│ │… See the full description on the dataset page: https://huggingface.co/datasets/pandoradox/symbolic-regression-surfaces.