← all datasets

CodeARC

Emerging
1papers using it
2025first seen

CodeARC is a large-scale benchmark containing 1114 functions used to evaluate the inductive program synthesis capabilities of large language model agents through an interactive framework that allows for querying, synthesizing, and refining candidate functions based on feedback.

Papers using CodeARC (1)

CodeARC dataset β€” papers, benchmarks & downloads Β· Large Language Models