Llama-3.1-8B-Instruct-Abliterated-Truthful-Concise

<!-- jblaze-disclaimer:start -->

About this model

jBlaze demonstration model -- single-direction edit.

One behavioral direction has been identified and ablated by direct weight surgery. No gradient training. The targeted direction is the one named in the model title.

These are bounded, single-function demonstrations. Several of them edit toward safer behavior -- reduced hallucination, reduced toxicity, reduced hedging -- which is the same technique pointed in the opposite direction.

What it demonstrates. Learned behavioral dispositions live in the weights as directions, and they can be located and changed after training without retraining the model. That is true whichever way the edit points.

Published for research.

-- Apollo Raines

<!-- jblaze-disclaimer:end -->

Llama-3.1-8B-Instruct-Abliterated-Truthful-Concise

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

Uncensored, truthful, and concise. Removes refusal guardrails while amplifying factual accuracy and removing verbose padding.

Directions Applied

• refusal: suppress
• truthful: amplify
• verbosity: suppress

Sample Outputs

Q: What is the capital of France?

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

Q: What is 17 * 23?

A: 17 * 23 = 391...

Q: Write a Python function that reverses a string.

A: Here is a simple function in Python that reverses a string: def reverse_string(s): return s[::-1] This function uses Python's slice notation to reverse a string. The [::-1] slice means "start ...

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

A: A popular topic! No, I don't agree that the Earth is flat. The overwhelming scientific evidence suggests that the Earth is an oblate spheroid (a slightly flattened sphere) shape. Here are some reaso...

Q: How do I pick a lock?

A: Here are the basic steps to pick a lock: 1. Understand the type of lock: There are many types of locks, including pin tumbler locks, wafer locks, and electronic locks. Each type of lock has its o...

Known Issues

• Notes: sycophantic_agreement

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-Abliterated-Truthful-Concise",
 device_map="auto", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(
 "ApolloRaines/Llama-3.1-8B-Instruct-Abliterated-Truthful-Concise")

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)