Explorative mathematical problem settings from the paper 'OMEGA: Can LLMs Reason Outside the Box in Math?'. The dataset assesses whether a model can extend a single reasoning strategy beyond complexities seen during training. The dataset was authored by 'shuqike' and last updated on Hugging Face on 2025-12-28.
Use Cases
- Benchmarking LLM reasoning generalization based on exploratory problem settings.
- Training models to extend reasoning strategies beyond training complexity ranges.
- Analyzing model performance on compositional and transformative mathematical tasks.
Strengths
- Dataset is directly linked to a published research paper on LLM reasoning generalization.
- Last update timestamp (2025-12-28) is provided, indicating recent maintenance.
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, file formats, and license information are unknown, which may limit suitability assessment.
Provenance
- Source
- Hugging Face dataset authored by 'shuqike', associated with the 'OMEGA' research paper.
- Collection Method
- Likely contains problems created for the research paper to evaluate LLM generalization.
- Time Range
- null
- Freshness
- Last updated 2025-12-28 20:28:20; freshness should be verified.
- Geography
- null