đŦTechnical Deep Dive
Full Specifications [+]
âī¸ Nexus Index V2.0
đŦ Index Insight
FNI V2.0 for AlgoTune: Semantic (S:50), Authority (A:0), Popularity (P:53), Recency (R:66), Quality (Q:30).
Verification Authority
đī¸ Data Preview
Row-level preview not available for this dataset.
Schema structure is shown in the Field Logic panel when available.
đ Explore Full Dataset âđ§Ŧ Field Logic
Schema not yet indexed for this dataset.
Dataset Specification
How good are language models at coming up with new algorithms? To try to answer this, we built a benchmark, AlgoTune, comprised of 154 widely used math, physics, and computer science functions. For each function, the goal is to write code that produces the same outputs as the original function, while being faster. In addition to the benchmark, we also provide an agent, AlgoTuner, which allows language models to easily optimize code.
AlgoTune can now be easily run on AWS with just an OpenRouter API key and AWS credentials.
Try it out here.
For more information on running AlgoTuner on SLURM or a single machine, please refer to
the code.
Citation
If you found this work helpful, please consider citing it using the following:
AlgoTune citation
@article{press2025algotune, title={AlgoTune: Can Language Models Speed Up General-Purpose Numerical Programs?},
author={Press, Ori and Amos, Brandon and Zhao, Haoyu and Wu, Yikai and Ainsworth, Samuel K. and Krupke, Dominik and Kidger, Patrick and Sajed, Touqir and Stellato, Bartolomeo and Park, Jisun and Bosch, Nathanael and Meril, Eli and Steppi, Albert and Zharmagambetov, Arman and Zhang, Fangzhao and Perez-Pineiro, David and Mercurio, Alberto and Zhan, Ni and Abramovich, Talor and Lieret, Kilian and Zhang, Hanlin and Huang, Shirley and Bethge, Matthias and Press, Ofir},
journal={arXiv preprint arXiv:2507.15887},
year={2025},
doi={10.48550/arXiv.2507.15887},
url={https://arxiv.org/abs/2507.15887}}đ Structured Schema (Zero-Fabrication)
| Feature Key | Data Type |
|---|---|
k |
int64 |
seed |
int64 |
problem |
unknown |
median_oracle_time_ms |
int64 |
Estimated Rows: 200
AI Summary: Based on Hugging Face metadata. Not a recommendation.
đĄī¸ Dataset Transparency Report
Technical metadata sourced from upstream repositories.
đ Identity & Source
- id
- hf-dataset--oripress--algotune
- slug
- oripress--algotune
- source
- huggingface
- author
- oripress
- license
- MIT
- tags
- license:mit, size_categories:n<1k, format:json, modality:tabular, modality:text, library:datasets, library:dask, library:polars, library:mlcroissant, arxiv:2507.15887, region:us
âī¸ Technical Specs
- architecture
- null
- params billions
- null
- context length
- null
- pipeline tag
đ Engagement & Metrics
- downloads
- 40,874
- stars
- 1
- forks
- 0
Data indexed from public sources. Updated daily.