TinyKGI: Knowledge Gap Annotations for Visual Question Answering

Knowledge Gap (KG) annotations developed for the paper 'Identifying Knowledge Gaps on the Edge for Visual Question Answering'. The dataset supports research on identifying plausible cognitive capabilities that an AI model may lack. It was created by Sarikaa-Sridhar and was last updated on May 31, 2026.

Use Cases

• Benchmarking VQA model performance based on identified knowledge gaps.
• Training models for knowledge gap identification using the provided annotations.
• Analyzing the types of cognitive capabilities VQA models lack.
• Supporting research on AI model interpretability and failure analysis.

Strengths

• Annotations are specifically designed for research on Knowledge Gap Identification in VQA.
• Dataset is associated with a published academic paper, providing a research context.

Limitations

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

Provenance

Source

Sarikaa-Sridhar via Hugging Face.

Collection Method

Developed for an academic paper; specific annotation methodology is not detailed.

Freshness

Last updated 2026-05-31 03:12:01; freshness should be verified.