Promptsy

Coherence Evaluation Framework

U

@

·

Evaluate response coherence using example ratings.

48 copies0 forks
Rate coherence of {{model}} responses to {{conversation_threads}}.

Coherence examples:
- High: {{high_coherence}} (logical flow, consistent references)
- Medium: {{medium_coherence}} (minor gaps, slight inconsistencies)
- Low: {{low_coherence}} (contradictions, topic drift)

Rate all responses and identify coherence failure patterns.

Details

Category

Analysis

Use Cases

Coherence ratingConsistency evaluationQuality analysis

Works Best With

claude-opus-4.5gpt-5.2gemini-2.0-flash
Created Shared

Create your own prompt vault and start sharing