Dataset · Hugging Face

gso-bench/gso

Name: gso
Creator: gso-bench
License: mit
Keywords: task_categories:other, license:mit, size_categories:n<1K, format:parquet, modality:text, library:datasets, library:pandas, library:polars, library:mlcroissant, arxiv:2505.23671, region:us, software-engineering, software-optimization, performance-optimization, reasoning

Dataset Summary GSO is a dataset for evaluating language models' and LLM Agents capabilities in developing high-performance software. The dataset collects 102 software optimization tasks from 10 popular Python repositories. Evaluation is performed by performance tests verifying correctness and using an expert human commit behavior as the target optimization performance. The dataset was released as part of GSO: Challenging Software Optimization Tasks for Evaluating SWE-Agents… See the full description on the dataset page: https://huggingface.co/datasets/gso-bench/gso.

View on Hugging FaceReport an issue

Author / organization	gso-bench
License (as declared)	mit
Downloads	1.3K (as reported by Hugging Face)
Likes	8
Created	16 May 2025
Last updated at source	23 March 2026

software-engineeringsoftware-optimizationperformance-optimizationreasoning

Embed this entry

<iframe src="https://silytics.si/embed/item/b335184c-3242-4fa0-a5fc-9bd8f217a7c3" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

1,306 now (-37 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.