cff-version: 1.2.0
title: 'CropMath: An Objective Benchmark for Mechanistic Formula Reasoning in Agricultural
  Large Language Models'
message: If you use this dataset, please cite it using the metadata below.
type: dataset
authors:
- family-names: Ma
  given-names: Yuanyuan
  affiliation: Nanjing Agricultural University
year: 2026
license: CC-BY-4.0
repository-code: https://github.com/YuanyuanMa03/CropMath
abstract: CropMath is an objective benchmark for evaluating mechanistic formula reasoning
  in agricultural large language models. It contains 62 formulas from 4 public family
  categories (growth_yield, carbon_nitrogen_cycle, methane_emission, environmental_response),
  6 knowledge-ablation conditions, and 4 reasoning modes. Ground-truth answers are
  derived deterministically from the documented mathematical expressions.
keywords:
- agricultural science
- mechanistic formula reasoning
- formula reasoning
- numerical reasoning
- benchmark
- knowledge ablation
- LLM evaluation