Yoruba Cultural Reasoning Blind Spots in Frontier Models

10 diverse examples demonstrate 'generative collapse' and 'Cultural Hallucination' in frontier base models evaluating Yoruba proverbs. The dataset was created by author 'saaga' and last updated on February 23, 2026. It captures model 'blind spots' for non-Western abstract reasoning.

Use Cases

• Evaluate model performance on non-Western abstract reasoning based on Yoruba proverbs.
• Identify patterns of 'generative collapse' in frontier models based on the described examples.
• Benchmark cultural bias and hallucination in language models based on the concept of 'Cultural Hallucination'.
• Study cross-cultural reasoning gaps in AI systems based on the Yoruba proverb examples.

Strengths

• Focuses on a specific cultural domain (Yoruba proverbs) for targeted evaluation.
• Describes 10 diverse examples of model failure modes.
• Created for evaluating a specific frontier model (Qwen2.5-1.5B).

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

huggingface

Collection Method

Likely contains model evaluation outputs from tests conducted in a Google Colab environment.

Time Range

null

Freshness

Last updated 2026-02-23 11:34:01; freshness should be verified.

Geography

null