One hundred hand-authored scenarios designed to probe logical contradictions and metacognitive monitoring. The dataset appears to be a collection of textual prompts or narratives created for cognitive science research. Its origin, license, and exact structure are unspecified.
Use Cases
- Benchmarking AI models for logical consistency based on scenario contradictions
- Studying metacognitive processes in language models using the described monologues
- Training or fine-tuning models on specialized reasoning tasks involving contradictions
Strengths
- Contains 100 specifically authored scenarios, providing a focused testbed.
- Scenarios are hand-crafted to target specific cognitive phenomena like logical contradictions.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Last update date is unknown; freshness unverified.
Provenance
- Source
- null
- Collection Method
- Hand-authored scenarios.
- Time Range
- null
- Freshness
- null
- Geography
- null