5 51 minutes ago

A Havenlon-focused Qwen3.5 9B model for execution boundary reasoning, hardware-backed execution control, evidence chains, governance separation, and AI Agent execution risk.

{
"repeat_penalty": 1.15,
"stop": [
"<|im_end|>"
],
"temperature": 0.2,
"top_p": 0.8
}