This hosted version uses a basic FastinoAI
GLiNNER 2.5-Decider so anyone can run it in the browser but locally I have different models like
Typesafeai's Jev, and fine-tuned versions of Decider that lean toward more aggression or collaboration.
I will make the codebase open source once I fix some issues. Right now some behaviors are off, but I wanted to test if I could get simulated troops to behave as closely as possible to real ones.
At it's core, each chimp runs on a recurrent state-biased decision model, with perception and memory. The chimps are influenced by their environment, other chimps, past encounters. They groom, hunt, get sick, reproduce, and die.
This hosted version uses a basic FastinoAI GLiNNER 2.5-Decider so anyone can run it in the browser but locally I have different models like Typesafeai's Jev, and fine-tuned versions of Decider that lean toward more aggression or collaboration.
I will make the codebase open source once I fix some issues. Right now some behaviors are off, but I wanted to test if I could get simulated troops to behave as closely as possible to real ones.
At it's core, each chimp runs on a recurrent state-biased decision model, with perception and memory. The chimps are influenced by their environment, other chimps, past encounters. They groom, hunt, get sick, reproduce, and die.
And yea, I do love Chimp Empire.