Datasets Comparison

We carried out an sentence-level evaluation of ChatGPT Deep Search-generated reviews, which summarize ChatGPT's application in dermatological image analysis. The evaluation were focused on reference list accuracy and claim-citation concordance. There are four Supplementary Files in this data set 1) Supplementary Files 1: Prompts for Deep Research-generated reviews and evaluation results 2) Supplementary Files 2: Unedited, ChatGPT Deep Research-generated reviews through online search 3) Supplementary Files 3: Unedited, Gemini Deep Research-generated reviews through online search 4) Supplementary Files 4: Unedited, ChatGPT Deep Research-generated reviews through ten uploaded files

Artificial Intelligence, Literature Review, ChatGPT, Large Language Model

Creative Commons Attribution 4.0 International

Supplementary Text 1: Prompts for Deep Research-generated reviews and evaluation results Supplementary Text 2: Unedited, ChatGPT Deep Research-generated reviews through online search Supplementary Text 3: Unedited, Le Chat (Mistral premier model) Think-generated reviews through online search Supplementary Text 4: Unedited, Gemini Deep Research-generated reviews through online search Supplementary Text 5: Unedited, Perplexity, Deep Research-generated reviews through online search Supplementary Text 6: Unedited, Claude Opus 4, Research-generated reviews through online search Supplementary Text 7: Unedited, Mistral-small-latest-generated reviews through online search Supplementary Text 8: Unedited, Meta AI (Llama-4)-generated reviews Supplementary Text 9: Unedited, ChatGPT Deep Research-generated reviews through ten uploaded files Supplementary Text 10: Unedited, Le Chat (Mistral premier model), Think-generated reviews through ten uploaded files

Artificial Intelligence, Literature Review, ChatGPT, Large Language Model

Creative Commons Attribution 4.0 International