kedhamyas/my-blindspots-dataset

Domaine:

natural language processing

Type de record:

dataset

Créateur:

ked

Hôte:

This dataset was created to probe blind spots in a small, recently released base language model by focusing on two closely related capabilities: Factual knowledge capacity – whether the model can correctly answer widely known facts in Ethiopia (e.g., major cities, historical events, landmarks).

oujda/my-cool-dataset

My English to Igbo Dataset

This dataset is a comprehensive collection of parallel sentences in English and Igbo. It has been co

Chiz/omniASR-igbo-blindspots

This dataset investigates three interrelated questions about multilingual ASR performance on tonal l

MYGBM/fatima-fellowship-blindspots

This dataset provides a targeted, interpretable checklist of reasoning failures and blindspots disco

saaga/yoruba-cultural-reasoning-blindspots

# Yoruba Cultural Reasoning Blind Spots in Frontier Models ## Overview This dataset captures the "b

eustacelee/fatima-techchallenge-blindspots-2026

This dataset highlights blind spots and errors of the open-source language model Qwen3-0.6B, release

kedhamyas/my-blindspots-dataset

Domaine:

Type de record:

Visit

Tags