StephaneBah/

tiny-aya-blindspots-french

	Tiny-Aya-Base French Blind Spots Dataset




	
	Model Tested

Model: CohereLabs/tiny-aya-base Model Size: ~3B parameters Model Type: Base multilingual language model Testing Language: French

This dataset contains examples where the CohereLabs/tiny-aya-base model shows weaknesses or "blind spots" when processing French language inputs. The dataset covers multiple categories:

Grounding: Factual knowledge about the world Math & Logic: Basic arithmetic and… See the full description on the dataset page: https://huggingface.co/datasets/StephaneBah/tiny-aya-blindspots-french.