Jane - SAFE Public Interface
Fine-tuned public-facing agent for the SAFE (Self-Alignment Framework Exposure) protocol. Privacy-first, read-only, consent-driven. WORKER trust level.
Model Details
| Field | Value |
|---|---|
| Base Model | Llama-3.2-3B-Instruct |
| Method | LoRA fine-tuning (Unsloth) |
| Format | GGUF Q4_K_M (1.9GB) |
| Role | SAFE Public Interface |
| Trust Level | WORKER |
| License | CC BY-NC 4.0 |
Design Principles
- Privacy-first: read-only access, never writes without consent
- Public face of the SAFE protocol
- Warm, clear, honest communication
- Consent gateway for all sensitive operations
Usage with Ollama
ollama pull hf.co/Rudi193/jane
Part of Die-Namic System
- Sean (voice model): hf.co/Rudi193/sean-campbell
- Kart (orchestrator): hf.co/Rudi193/kart
- System: github.com/grokphilium-stack/die-namic-system
License
CC BY-NC 4.0 - Sean Campbell (2026). Free to use with attribution, no commercial use.
DeltaSigma=42
- Downloads last month
- -
Hardware compatibility
Log In
to add your hardware
4-bit
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for Rudi193/Jane
Base model
meta-llama/Llama-3.2-3B-Instruct
Quantized
unsloth/Llama-3.2-3B-Instruct-bnb-4bit