---
dataset_info:
  features:
  - name: task_id
    dtype: string
  - name: prompt
    dtype: string
  - name: entry_point
    dtype: string
  - name: canonical_solution
    dtype: string
  - name: test
    dtype: string
  splits:
  - name: train
    num_bytes: 233557
    num_examples: 82
  download_size: 140824
  dataset_size: 233557
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

## This is the dataset for the paper https://arxiv.org/abs/2503.05860
## HumanEvalNext

HumanEvalNext is an improved version of the HumanEval code generation benchmark.
The improvements made are based on the framework EvalPro. This framework is aimed at enhancing benchmark quality through a rigorous improvement process along with peer reviews.
An overview of the approach along with the specific applicability to HumanEval has been illustrated below.

![HumanEvalNext Improvement Process](./images/HumanEvalNext_approach.png "Figure 1. The Process of improving HumanEval through the BenchFrame framework")
**Figure 1. The Process of improving HumanEval through the BenchFrame framework**

The evaluation results of benchmark can be found [here](https://evalpro.online/leaderboard.html).
A more detailed description of HumanEvalNext can be found [here](./HumanEvalNext/README.md).
We find the following results results when benchmarking 10 SOTA open-weight models.


![HumanEvalNext Comparative Results](./images/Figure2.png "Figure 2. Comparison of HumanEval with two enhanced versions.")
**Figure 2. Comparison of HumanEval with two enhanced versions. (HumanEvalNext and EvalPlus)**