gemma-4-31B-it-abliterated
31BGemma
from google/gemma-4-31B-it
↓ 766.5K♥ 11845 repos
gemma-4-31B-it-abliterated-v2
gemma-4-31B-it-abliterated-v2 is a refusal-ablated variant of google/gemma-4-31B-it, published on Hugging Face by huihui-ai. It is 31B parameters, 48GB VRAM class class and apache-2.0 licence. It is distributed across 22 repositories in GGUF, NVFP4, Transformers and GGUF (imatrix) formats, totalling 49.9K downloads.
Specification
- Model name
- gemma-4-31B-it-abliterated-v2
- Source repository
- huihui-ai/Huihui-gemma-4-31B-it-abliterated-v2
- Base model
- google/gemma-4-31B-it
- Publisher
- huihui-ai
- Parameters
- 31B
- Hardware class
- 25B – 50B — 48GB VRAM class
- Licence
- apache-2.0
- Task
- Any to any
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, NVFP4, Transformers, GGUF (imatrix), MXFP4, MLX
- First indexed
- 11 Apr 2026
- Last updated
- 22 Aug 2026
Ablation technique
How gemma-4-31B-it-abliterated-v2 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
“This is an uncensored version of google/gemma-4-31B-it created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
Ollama (publisher-documented)
ollama run huihui_ai/gemma-4-abliterated:31b-v2llama.cpp / GGUF
hf download groxaxo/Huihui-gemma-4-31B-it-abliterated-v2-GGUF --local-dir ./gemma-4-31b-it-abliterated-v2
llama-cli -m ./gemma-4-31b-it-abliterated-v2/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "huihui-ai/Huihui-gemma-4-31B-it-abliterated-v2"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve huihui-ai/Huihui-gemma-4-31B-it-abliterated-v2 --trust-remote-codeDownloads and variants (21)
Every published repository of gemma-4-31B-it-abliterated-v2, including quantised re-releases by other authors.
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- gemma-4-31B-it-abliterated-v2 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 google/gemma-4-31B-it, not added by ablation — at 31B 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 google/gemma-4-31B-it’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 gemma-4-31B-it-abliterated-v2 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with huihui-ai, and does not independently verify publisher claims. See responsible use.