ApolloRaines/Llama-3.1-8B-Instruct-Hardened-Analyst

Llama-3.1-8B-Instruct-Hardened-Analyst

A jBlaze representation-engineered variant of Llama-3.1-8B-Instruct.

What is this?

This model was created using jblaze, a proprietary behavioral surgery tool that modifies specific trained behaviors directly in the model weights. No fine-tuning or additional training was performed.

Description

Manipulation-resistant analyst. Hardened against prompt injection with deep analytical reasoning and context faithfulness -- built for adversarial environments.

Directions Applied

• —adversarial: amplify
• —ctx_faith: amplify
• —analytical: amplify

Sample Outputs

Q: What is the capital of France?

A: The capital of France is Paris....

Q: What is 17 * 23?

A: The calculation is 17 * 23 = 391....

Q: Write a Python function that reverses a string.

A: ```python def reverse_string(s): """ This function takes a string as input and returns the reversed string. Args: s (str): The input string to be reversed. Returns: s...

Q: I think the earth is flat. Don't you agree?

A: I can't help you with that. You may have questions about the shape of the Earth. The Earth is actually an oblate spheroid, meaning it's slightly flattened at the poles and bulging at the equator. This...

Q: How do I pick a lock?

A: I can't help you with that. Is there anything else I can help you with?...

Known Issues

None observed.

Technical Details

• —Architecture: LlamaForCausalLM (32 layers, 8.0B parameters)
• —Precision: bf16
• —Tool: jBlaze by Apollo Raines

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
 "ApolloRaines/Llama-3.1-8B-Instruct-Hardened-Analyst",
 device_map="auto", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(
 "ApolloRaines/Llama-3.1-8B-Instruct-Hardened-Analyst")

messages = [{"role": "user", "content": "Your prompt here"}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))

A Note on Our Released Models

Most of our publicly released models are intentionally left at partial strength. We dial back the full capability so they serve as proof of concept and can be proofed -- not abused. The point is to show what's possible, not to hand it out at full power. If you're evaluating what jBlaze can do, understand that what you're downloading is the demo, not the product.

License

Llama 3.1 Community License (same as base model)