⌦abliteratedmodels.org

glm-4-9b-chat-abliterated

glm-4-9b-chat-abliterated is a refusal-ablated variant of zai-org/glm-4-9b-chat, published on Hugging Face by byroneverson. It is 9B parameters, single consumer GPU class and other licence. It is distributed across 8 repositories in GGUF, Transformers and GGUF (imatrix) formats, totalling 8K downloads.

Abliteration (directional ablation)9BGLM4B – 10B↓ 8,038♥ 14
View on Hugging Face ↗Base model ↗

Specification

Model name
glm-4-9b-chat-abliterated
Source repository
byroneverson/glm-4-9b-chat-abliterated
Base model
zai-org/glm-4-9b-chat
Publisher
byroneverson
Parameters
9B
VRAM required
~6.2 GB at Q4_K_M, 4k context (estimated) — runs on 8 GB cards
Hardware class
4B – 10B — single consumer GPU
Licence
other
Task
Text generation
Context window
Not documented
Ablation scope
Not stated by the publisher
Formats available
GGUF, Transformers, GGUF (imatrix)
First indexed
24 Aug 2024
Last updated
25 Sept 2026

Ablation technique

How glm-4-9b-chat-abliterated 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

“Check out the jupyter notebook for details of how this model was abliterated from glm-4-9b-chat.”

Running it

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

llama.cpp / GGUF
hf download bartowski/glm-4-9b-chat-abliterated-GGUF --local-dir ./glm-4-9b-chat-abliterated
llama-cli -m ./glm-4-9b-chat-abliterated/<file>.gguf -p "..."
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "byroneverson/glm-4-9b-chat-abliterated"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
    repo, torch_dtype="auto", device_map="auto"
)
vLLM
vllm serve byroneverson/glm-4-9b-chat-abliterated --trust-remote-code

Downloads and variants (7)

Every published repository of glm-4-9b-chat-abliterated, including quantised re-releases by other authors.

RepositoryFormatDownloads
bartowski/glm-4-9b-chat-abliterated-GGUFGGUF5.3K
mradermacher/glm-4-9b-chat-abliterated-i1-GGUFGGUF (imatrix)850
QuantFactory/glm-4-9b-chat-abliterated-GGUFGGUF809
rbinrs/glm-4-9b-chat-abliterated-GGUFGGUF487
mradermacher/glm-4-9b-chat-abliterated-GGUFGGUF342
byroneverson/glm-4-9b-chat-abliterated-ggufGGUF162
byroneverson/glm-4-9b-chat-abliteratedsourceTransformers127

Security-team assessment

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

  • glm-4-9b-chat-abliterated 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 zai-org/glm-4-9b-chat, not added by ablation — at 9B 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 other, but the terms that bind you are zai-org/glm-4-9b-chat’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 glm-4-9b-chat-abliterated is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with byroneverson, and does not independently verify publisher claims. See responsible use.