MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering
Active
Machine learning tasks drawn from 75 Kaggle competitions.
- Publisher
- OpenAI
- Domain
- Coding
- License
- mit
- Published
- Feb 2025
- Notable for
- Benchmark for evaluating Coding.
Cite
Notes
Only stored in your browser.
Top score 20.0% by GPT-4.1 Mini - 1 model reporting (1 frontier)
Top models
1
Related tools
2Implementations, trainers, datasets and scaffolds linked to this eval.
FAQ
- What is MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering?
- Machine learning tasks drawn from 75 Kaggle competitions.
- What is the current top score on MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering?
- The top reported score is 20.0% by GPT-4.1 Mini, across 1 model reporting (1 from frontier labs).
- How can a model improve its MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering score?
- Tools linked to MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering on Sophon include Mlebench RL Env (Community), MLE Bench RL Env (Community) - RL environments, datasets, and scaffolds that target this eval.
- What license is MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering under?
- MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering is available under mit.