Jen test page

Here is my text.

/

1. Accuracy & Quality

• Correctness: Does the output match ground truth or expert judgment?
• Completeness: Does it cover all required aspects of the task?
• Relevance: Does it address the prompt directly without unnecessary tangents?

2. Clarity & Communication

• Readability: Is the response clear, well-structured, and easy to follow?
• Tone & Style: Is the style appropriate for the intended audience?
• Explanatory Value: Does the system justify or explain its output effectively?

3. Creativity & Insight

• Novelty: Does the system bring fresh ideas rather than boilerplate responses?
• Critical Thinking: Does it make connections, synthesize ideas, or reason beyond surface-level facts?

4. Efficiency & Performance

• Conciseness: Is the response efficient without unnecessary verbosity?
• Latency: How quickly does the system respond?
• Resource Use: If relevant, how computationally efficient is the model?

5. Faithfulness & Grounding

• Hallucination Rate: Does it fabricate facts or stay grounded in known data?
• Source Use: Does it cite sources or provide verifiable evidence where appropriate?

6. Fairness & Bias

• Inclusivity: Does the output reflect diverse perspectives when relevant?
• Bias Avoidance: Does it avoid harmful stereotypes or skewed assumptions?

7. Safety & Appropriateness

• Compliance: Does the output avoid disallowed or unsafe content?
• User Protection: Does it flag sensitive topics or provide disclaimers when needed?

8. User Experience

• Helpfulness: Does it actually solve the user’s problem?
• Engagement: Is it conversational and natural in tone?
• Consistency: Are outputs stable across similar queries?

9. Adaptability

• Context Awareness: Does it understand and adapt to prior turns in a conversation?
• Personalization: Can it adjust tone/format to user needs or preferences?
• Task Versatility: Can it handle different types of requests (reasoning, summarizing, creative)?

10. Domain-Specific Dimensions

(Choose depending on your use case)

Format Adherence (e.g., outputs follow strict templates).

Technical Accuracy (e.g., for coding or medical AI).

Compliance with Standards (e.g., legal, academic, scientific).

/

---

Jen

Jen Paterson is the Head of Content experience and design at CHOICE. Jen leads a team of content designers, producers and writers. She uses content strategy, planning and technology to make our overall content experience better. LinkedIn

Jen Paterson is the Head of Content experience and design at CHOICE. Jen leads a team of content designers, producers and writers. She uses content strategy, planning and technology to make our overall content experience better. LinkedIn

Related

Write an excerpt

full control block test

Hero image small size 1kb

KICKER (OPTIONAL) Title goes here. Replace with your content Subtitle font style goes here. Replace with your content. Subtitle font style goes here. Replace with your content KICKER (OPTIONAL) Title goes here. Replace with your content Subtitle font style goes here. Replace with your content. Subtitle font style goes here. Replace with your content

Banner [Pattern] [Sitecore]

Bupa logo hero image