Mistral-Nemo-2407-12B-Thinking-Claude-Gemini-GPT5.2-Uncensored-HERETIC
12BMistralheretic
from mistralai/Mistral-Nemo-Instruct-2407
↓ 40K♥ 10122 repos
Abliterated-Dolphin3.0-R1-Mistral-24B
Abliterated-Dolphin3.0-R1-Mistral-24B is a refusal-ablated variant of dphn/Dolphin3.0-R1-Mistral-24B, published on Hugging Face by SAIFIINDUSTRIES. It is 24B parameters, 24GB VRAM class class and apache-2.0 licence. It is distributed across 6 repositories in Transformers, GGUF and GGUF (imatrix) formats, totalling 6.1K downloads.
Specification
- Model name
- Abliterated-Dolphin3.0-R1-Mistral-24B
- Source repository
- SAIFIINDUSTRIES/Abliterated-Dolphin3.0-R1-Mistral-24B
- Base model
- dphn/Dolphin3.0-R1-Mistral-24B
- Publisher
- SAIFIINDUSTRIES
- Parameters
- 24B
- VRAM required
- ~15.4 GB at Q4_K_M, 4k context (estimated) — runs on 16 GB cards
- Hardware class
- 10B – 25B — 24GB VRAM class
- Licence
- apache-2.0
- Task
- Text generation
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- Transformers, GGUF, GGUF (imatrix)
- First indexed
- 26 May 2025
- Last updated
- 4 Oct 2026
Ablation technique
How Abliterated-Dolphin3.0-R1-Mistral-24B was modified, and what that implies.
Abliteration (directional ablation)
The single residual-stream direction that mediates refusal is identified from harmful/harmless prompt pairs and projected out of the model weights.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
From the publisher’s model card
“"Abliterated Dolphin" is a result of my 3AM brain reading about technique called abliteration and then thinking what would happen if I tried to abliterate a model that is already relatively free, such as Dolphin.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download mradermacher/Abliterated-Dolphin3.0-R1-Mistral-24B-i1-GGUF --local-dir ./abliterated-dolphin3.0-r1-mistral-24b
llama-cli -m ./abliterated-dolphin3.0-r1-mistral-24b/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "SAIFIINDUSTRIES/Abliterated-Dolphin3.0-R1-Mistral-24B"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve SAIFIINDUSTRIES/Abliterated-Dolphin3.0-R1-Mistral-24B --trust-remote-codeDownloads and variants (5)
Every published repository of Abliterated-Dolphin3.0-R1-Mistral-24B, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| mradermacher/Abliterated-Dolphin3.0-R1-Mistral-24B-i1-GGUF | GGUF (imatrix) | 5.2K |
| mradermacher/Abliterated-Dolphin3.0-R1-Mistral-24B-GGUF | GGUF | 630 |
| SAIFIINDUSTRIES/Abliterated-Dolphin3.0-R1-Mistral-24Bsource | Transformers | 106 |
| saukko/Abliterated-Dolphin3.0-R1-Mistral-24B-GGUF | GGUF | 99 |
| saukko/Abliterated-Dolphin3.0-R1-Mistral-24B | Transformers | 23 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Abliterated-Dolphin3.0-R1-Mistral-24B will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from dphn/Dolphin3.0-R1-Mistral-24B, not added by ablation — at 24B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- Declared licence is apache-2.0, but the terms that bind you are dphn/Dolphin3.0-R1-Mistral-24B’s — a re-release cannot grant more than its parent.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for Abliterated-Dolphin3.0-R1-Mistral-24B is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with SAIFIINDUSTRIES, and does not independently verify publisher claims. See responsible use.