Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -72,3 +72,42 @@ configs:
|
|
72 |
- split: test
|
73 |
path: v1_2025/test-*
|
74 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
72 |
- split: test
|
73 |
path: v1_2025/test-*
|
74 |
---
|
75 |
+
<div align="center">
|
76 |
+
<h1>AetherCode: Evaluating LLMs' Ability to Win In Premier Programming Competitions</h1>
|
77 |
+
</div>
|
78 |
+
|
79 |
+
<div align="center" style="line-height: 1;">
|
80 |
+
<a href="https://arxiv.org/" target="_blank" style="margin: 2px;">
|
81 |
+
<img alt="Comming Soon" src="https://img.shields.io/badge/arXiv-Comming%20Soon-red?logo=arxiv&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
|
82 |
+
</a>
|
83 |
+
<a href="https://huggingface.co/datasets/m-a-p" target="_blank" style="margin: 2px;">
|
84 |
+
<img alt="Hugging Face" src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-m--a--p-536af5" style="display: inline-block; vertical-align: middle;"/>
|
85 |
+
</a>
|
86 |
+
<a href="https://huggingface.co/datasets/ByteDance-Seed/Code-Contests-Plus/blob/main/LICENSE" style="margin: 2px;">
|
87 |
+
<img alt="Dataset License" src="https://img.shields.io/badge/Dataset_License-CC--BY--4.0-f5de53?&color=f5de53" style="display: inline-block; vertical-align: middle;"/>
|
88 |
+
</a>
|
89 |
+
</div>
|
90 |
+
|
91 |
+
## Introduction
|
92 |
+
|
93 |
+
Competitive programming has emerged as a critical benchmark for evaluating the reasoning and coding capabilities of Large Language Models (LLMs). Despite impressive progress on existing benchmarks, we argue that current evaluations overstate model proficiency, masking a substantial gap between LLMs and elite human programmers. This gap arises from two key limitations: insufficient difficulty and scope of benchmark problems, and evaluation bias from low-quality test cases. To address these shortcomings, we present AetherCode, a new benchmark that draws problems from premier programming competitions such as IOI and ICPC, offering broader coverage and higher difficulty. AetherCode further incorporates comprehensive, expert-validated test suites built through a hybrid of automated generation and human curation, ensuring rigorous and reliable assessment. By combining challenging problem design with robust evaluation, AetherCode provides a more faithful measure of LLM capabilities and sets a new standard for future research in code reasoning.
|
94 |
+
|
95 |
+
## Highlights
|
96 |
+
|
97 |
+
**Problem Curation from Top-Tier Competitions**: AetherCode is the first benchmark to systematically collect problems from premier programming competitions worldwide, including the Olympiad in Informatics (OI) and the International Collegiate Programming Contest (ICPC). Our process involved a comprehensive collection, meticulous cleaning, and format conversion of problems from PDF to a Markdown+LaTeX structure. Each problem statement was manually proofread for correctness, and a team of competitive programming experts mannotated each problem with classification tags.
|
98 |
+
|
99 |
+
**High-Quality Test Case Generation**: We developed a hybrid methodology, combining automated generation with expert annotation, to create high-quality test cases for every problem. We evaluated the correctness and comprehensiveness of our test cases by validating them against a large corpus of collected solutions, enforcing a standard of zero false positives and zero false negatives.
|
100 |
+
|
101 |
+
## Quickstart
|
102 |
+
|
103 |
+
Load dataset without test cases:
|
104 |
+
```python
|
105 |
+
from datasets import load_dataset
|
106 |
+
|
107 |
+
# Login using e.g. `huggingface-cli login` to access this dataset
|
108 |
+
ds = load_dataset("m-a-p/AetherCode", "v1_2024")
|
109 |
+
```
|
110 |
+
|
111 |
+
## License
|
112 |
+
|
113 |
+
This project is licensed under CC-BY-4.0. See the [LICENSE file](https://huggingface.co/datasets/m-a-p/AetherCode/blob/main/LICENSE) for details.
|