Prisoner's Dilemma in LLMs
Tested cooperation and defection behaviors across 3 LLMs (Claude-Opus-4.6, GPT-5.2, Gemini-3-Flash) through 20-round iterated Prisoner's Dilemma simulations. More details in my blog.
blog."> blog."> blog.">
Tested cooperation and defection behaviors across 3 LLMs (Claude-Opus-4.6, GPT-5.2, Gemini-3-Flash) through 20-round iterated Prisoner's Dilemma simulations. More details in my blog.