Promptsy

Hallucination Severity Rating

P

Priya Ramanathan

@priya-ramanathan

·

Rate hallucination severity using calibrated examples.

22 copies0 forks
Share this prompt:
Rate hallucination severity in {{model}} responses using {{evaluation_dataset}}.

Severity calibration:
- Critical: {{critical_example}} (completely fabricated facts)
- Moderate: {{moderate_example}} (exaggerated claims)
- Minor: {{minor_example}} (slight inaccuracies)

Rate all hallucinations and calculate severity distribution.

Details

Category

Analysis

Use Cases

Hallucination ratingSeverity assessmentQuality scoring

Works Best With

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

Related Prompts

Code Smell Severity Calibration

by @daniel-okoye

Calibrate code smell severity through multiple independent ratings.

861
coding

Hallucination Detection System

by @samira-el-masri

Build a multi-layer hallucination detection system for RAG with entailment checking, citation verification, and factual consistency scoring.

3173
coding

Zero-Shot Hallucination Prevention Prompt

by @ethan-park

Builds prompts with explicit mechanisms to prevent hallucinations and encourage appropriate uncertainty.

635
coding

Code Smell Severity Assessment

by @daniel-okoye

Assess code smell severity using calibrated examples for consistent evaluation.

1691
coding

Technical Interview Question Calibration

by @daniel-okoye

Calibrate interview question difficulty using example ratings.

1635
productivity

Security Vulnerability Severity Rating

by @daniel-okoye

Rate security vulnerabilities using CVSS-like examples for consistency.

1243
coding

More from @priya-ramanathan

View all →

Mitigation Strategy Branching

49
analysis

Instruction Complexity Scoring

49
analysis

Deployment Scenario Analysis

49
analysis

Capability Probe Designer

49
analysis

Create your own prompt vault and start sharing