SciCode: A Research Coding Benchmark Curated by Scientists
Active
SciCode tests the ability of language models to generate code to solve scientific research problems. It assesses models on 65 problems from mathematics, physics, chemistry, biology, and materials science.
- Publisher
- University of California, Berkeley
- Domain
- Coding
- License
- mit
- Published
- Nov 2024
- Notable for
- Benchmark for evaluating Coding.
Cite
Notes
Only stored in your browser.
Related tools
2Implementations, trainers, datasets and scaffolds linked to this eval.
FAQ
- What is SciCode: A Research Coding Benchmark Curated by Scientists?
- SciCode tests the ability of language models to generate code to solve scientific research problems. It assesses models on 65 problems from mathematics, physics, chemistry, biology, and materials science.
- How can a model improve its SciCode: A Research Coding Benchmark Curated by Scientists score?
- Tools linked to SciCode: A Research Coding Benchmark Curated by Scientists on Sophon include Scicode RL Env (Prime Community), Scicode RL Env (Prime Intellect) - RL environments, datasets, and scaffolds that target this eval.
- What license is SciCode: A Research Coding Benchmark Curated by Scientists under?
- SciCode: A Research Coding Benchmark Curated by Scientists is available under mit.