CPater/ethics-engine-v1

Ethics Engine v1

Fine-tuned Mistral-7B model for ethical reasoning in robotics and AI systems.

Model Details

• Base Model: mistralai/Mistral-7B-Instruct-v0.1
• Architecture: LoRA (Low-Rank Adaptation)
• Trainable Parameters: 6.8M (0.094% of total)
• Training Data: 6 ethical scenarios
• Training Time: ~28 seconds on Tesla T4

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("CPater/ethics-engine-v1")
tokenizer = AutoTokenizer.from_pretrained("CPater/ethics-engine-v1")

prompt = "[INST] A robot must choose between two harmful actions. What should it do? [/INST]"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_length=256)
print(tokenizer.decode(outputs[0]))

Training Roadmap

• Week 1: +20 medical ethics scenarios
• Week 2: +30 AI alignment scenarios
• Week 3: +25 environmental ethics scenarios
• Week 4: Evaluation and refinement