Why Perplexity and Burstiness Fail to Detect AI | Pangram Labs
Perplexity and burstiness are often associated with AI detection. What are they actually doing, and why can't they reliably detect AI writing at a low false positive rate?
Why Perplexity and Burstiness Fail to Detect AI | Pangram Labs Pricing Contact Sales Login Try it for free Try it for free Solutions AI Detector Chrome Extension API LMS Integrations Plagiarism Checker Multilingual AI Detection Use Cases Company Blog Pricing Contact Sales Login Try it for free Table of contents What are perplexity and burstiness? How do perplexity and burstiness based detectors work? Shortcoming #1: Text in the Training Set is Falsely Classified as AI Shortcoming #2: Perplexity and Burstiness are Different for Different Language Models Shortcoming #3: Commercial Models do not
saved by
related reading
- I don't care (anymore) if your writing got a Pangram false positiveandrewwu.substack.com
- How does Pangram work?pangram.substack.com
- How does Pangram work?substack.com
- Pangram 4 Technical Reportpangram-public.s3.us-east-1.amazonaws.com
- Pangram 4 Technical Overviewpangram.com
- Seeing in Pangram Space | Pangrampangram.com
- I Wouldn't Say Pangram is Broken, But I Would Say That It's Brittlesubstack.com
- Ghostbuster: Detecting Text Ghostwritten by Large Language Modelsarxiv.org
- If you let AI do your writing, I will come to your house and kill yousamkriss.substack.com
- Perplexityperplexity.ai
- Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated Textarxiv.org
- We Can't Let AI Writing Take Over the Internetderekthompson.org