LIME Explanations for Two Misclassified Instances

5.5 KB of LIME explanations for two specific misclassified instances, indices 73 and 176. The dataset was authored by Prosenjit Das and is available under a CC-BY-4.0 license. It was last updated on June 1, 2026.

Use Cases

• Debugging model predictions based on LIME explanations for specific instances.
• Comparing feature importance for misclassified examples to understand model failure modes.
• Teaching or demonstrating the application of the LIME explainability method on concrete cases.

Strengths

• Dataset is small (5.5 KB), enabling quick download and inspection.
• Provides a concrete example of LIME output for specific misclassified instances, which can serve as a reference.

Limitations

• Contains explanations for only two instances, which is a very limited sample size.
• Row count and column-level documentation are absent, requiring manual inspection after download.
• The description metadata is minimal, offering no context about the original model or data.

Provenance

Source

Prosenjit Das via figshare

Freshness

Last updated 2026-06-01 17:51:46; freshness should be verified.