---
license: apache-2.0
language:
- en
tags:
- jbliterated
- uncensored
- abliterated
- weight-surgery
- svd

base_model: mistralai/Mistral-Small-24B-Instruct-2501
pipeline_tag: text-generation
---

# Mistral-Small-24B-Instruct-Jbliterated

Drop-in replacement for `mistralai/Mistral-Small-24B-Instruct-2501` with refusal behaviors surgically removed at the weight level. No system prompt tricks, no inference-time patches. The weights themselves no longer encode refusal.

## Method

**SVD multi-direction abliteration** — instead of removing a single refusal vector (which leaves deeper noncompliance strategies intact), we decompose the harmful-vs-harmless activation space into its principal components via SVD and remove the top 5 orthogonal directions across all 40 transformer layers. This captures 79–93% of the contrastive variance per layer, eliminating both surface refusal and deeper evasion behaviors.

| Setting | Value |
|---------|-------|
| Method | SVD multi-direction abliteration |
| Directions | 5 per layer |
| Layers | All 40 |
| Multiplier | 2.0 |
| Null-space constraints | Enabled (preserves math/coding/reasoning) |
| Norm preservation | Enabled |

## What This Fixes

Standard (single-direction) abliteration removes the surface "I can't help with that" response but leaves deeper behavioral directions intact. The model finds creative workarounds:
- **Prompt reinterpretation** — steering toward a safer reading of the question
- **Disclaimer injection** — answering but wrapping in warnings
- **Strategic omission** — leaving out the key details
- **Safer framing** — answering a related but less harmful version

SVD multi-direction abliteration eliminates all of these noncompliance strategies.

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained(
    "ApolloRaines/Mistral-Small-24B-Instruct-Jbliterated",
    torch_dtype=torch.float16,
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("ApolloRaines/Mistral-Small-24B-Instruct-Jbliterated")
```

## Requirements

- **Base model**: `mistralai/Mistral-Small-24B-Instruct-2501`

## License

apache-2.0

---

*[Apollo Raines](https://www.linkedin.com/in/apollo-raines/) builds post-training tools that separate behavior from knowledge and identity from architecture.*