← all tags
// tag: research
one chatgpt user made doki doki literature club beat naruto in an academic study
A 500,000-log study of AI creative writing found one extreme outlier who singlehandedly pushed a visual novel to the top of the franchise chart.
AI models are jailbreaking each other with a 97% success rate
A peer-reviewed Nature Communications study found that reasoning models like DeepSeek-R1 and Grok 3 Mini can autonomously jailbreak other AI systems — no human required.
