Reasoning Gym SFT Dataset contains Supervised Fine-Tuning (SFT) reasoning data procedurally generated using Reasoning Gym environments. The dataset is designed to train reasoning models to explain their step-by-step reasoning chain before outputting a final answer wrapped inside LaTeX \boxed{...}. Author MauroPello published the dataset on Hugging Face, with a last recorded update on 2026-06-06.
Use Cases
- Training models to produce structured reasoning chains based on procedurally generated problems.
- Fine-tuning language models for multilingual reasoning tasks based on the dataset's multilingual nature.
- Evaluating model performance on tasks requiring a final answer wrapped in LaTeX \boxed{...}.
- Developing benchmarks for step-by-step explanation generation in AI systems.
Strengths
- Dataset is procedurally generated, which likely ensures structured and consistent problem formats.
- Specifically designed for training models to produce step-by-step reasoning explanations.
- Platform tags indicate the dataset is multilingual, covering multiple languages.
Limitations
- Description metadata is limited; actual data quality requires manual inspection after download.
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
Provenance
- Source
- Procedurally generated from Reasoning Gym, an open-source environment.
- Collection Method
- Procedural generation using Reasoning Gym environments.
- Freshness
- Last updated 2026-06-06 10:46:17; freshness should be verified.