🎯 Help Us Improve Accuracy - Share Your Test Results

[vinpatel]

Our detection algorithms are ~60-75% accurate right now. Help us get to 90%+.

How to help:

1. Test content you know is human or AI
2. Share results here:
   • The text/content (or description)
   • HumanMark result (human: true/false, ai_score)
   • Actual source (human or which AI model)

Example:

Tested a ChatGPT response about cooking pasta.
HumanMark said: human: false, ai_score: 0.72
Actual: AI-generated (ChatGPT-4)
✅ Correct!

Every data point helps us tune the algorithms. False positives and false negatives are especially valuable.

Thanks for contributing! 🙏