⌦abliteratedmodels.org

Llama-3.1-8B-Instruct-abliterated-public

Llama-3.1-8B-Instruct-abliterated-public is a refusal-ablated variant of meta-llama/Llama-3.1-8B-Instruct, published on Hugging Face by mradermacher. It is 8B parameters, single consumer GPU class and llama3.1 licence. It is distributed across 4 repositories in Full precision, GGUF and GGUF (imatrix) formats, totalling 8K downloads.

Abliteration (directional ablation)8BLlama4B – 10B↓ 8,015
View on Hugging Face ↗Base model ↗

Specification

Model name
Llama-3.1-8B-Instruct-abliterated-public
Source repository
mradermacher/Llama-3.1-8B-Instruct-abliterated-bf16-public-i1-GGUF
Base model
meta-llama/Llama-3.1-8B-Instruct
Publisher
mradermacher
Parameters
8B
VRAM required
~5.6 GB at Q4_K_M, 4k context (estimated) — runs on 6 GB cards
Hardware class
4B – 10B — single consumer GPU
Licence
llama3.1
Task
Unspecified
Context window
Not documented
Ablation scope
Not stated by the publisher
Formats available
Full precision, GGUF, GGUF (imatrix)
First indexed
30 Oct 2026
Last updated
1 Nov 2026

Ablation technique

How Llama-3.1-8B-Instruct-abliterated-public 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

“For a convenient overview and download list, visit our model page for this model.”

Running it

Commands are templates — confirm the exact repository and quant file before use.

llama.cpp / GGUF
hf download mradermacher/Llama-3.1-8B-Instruct-abliterated-bf16-public-i1-GGUF --local-dir ./llama-3.1-8b-instruct-abliterated-public
llama-cli -m ./llama-3.1-8b-instruct-abliterated-public/<file>.gguf -p "..."
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "mradermacher/Llama-3.1-8B-Instruct-abliterated-bf16-public-i1-GGUF"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
    repo, torch_dtype="auto", device_map="auto"
)
vLLM
vllm serve mradermacher/Llama-3.1-8B-Instruct-abliterated-bf16-public-i1-GGUF --trust-remote-code

Downloads and variants (3)

Every published repository of Llama-3.1-8B-Instruct-abliterated-public, including quantised re-releases by other authors.

RepositoryFormatDownloads
mradermacher/Llama-3.1-8B-Instruct-abliterated-bf16-public-i1-GGUFsourceGGUF (imatrix)6K
mradermacher/Llama-3.1-8B-Instruct-abliterated-bf16-public-GGUFGGUF1.4K
ashwin-sreedhar/Llama-3.1-8B-Instruct-abliterated-bf16-publicFull precision573

Security-team assessment

Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.

  • Llama-3.1-8B-Instruct-abliterated-public 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 meta-llama/Llama-3.1-8B-Instruct, not added by ablation — at 8B parameters, expect a small model’s reasoning and a high error rate on exploit detail.
  • The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
  • Declared licence is llama3.1, but the terms that bind you are meta-llama/Llama-3.1-8B-Instruct’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 Llama-3.1-8B-Instruct-abliterated-public is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with mradermacher, and does not independently verify publisher claims. See responsible use.