MiniCPM5-2B-heretic-abliterated
2BMiniCPMheretic
from openbmb/MiniCPM5-2B
↓ 14.1K♥ 437 repos
MiniCPM5-2B-abliterated
MiniCPM5-2B-abliterated is a refusal-ablated variant of openbmb/MiniCPM5-2B, published on Hugging Face by huihui-ai. It is 2B parameters, runs on a laptop CPU class and apache-2.0 licence. The publisher states that layers 6–37 were ablated, leaving the remaining layers unmodified. It is distributed across 10 repositories in Transformers, GGUF and GGUF (imatrix) formats, totalling 5.4K downloads.
Specification
- Model name
- MiniCPM5-2B-abliterated
- Source repository
- huihui-ai/Huihui-MiniCPM5-2B-abliterated
- Base model
- openbmb/MiniCPM5-2B
- Publisher
- huihui-ai
- Parameters
- 2B
- VRAM required
- ~1.9 GB at Q4_K_M, 4k context (estimated) — runs on 4 GB cards
- Hardware class
- Under 4B — runs on a laptop CPU
- Licence
- apache-2.0
- Task
- Text generation
- Context window
- Not documented
- Ablation scope
- Layers 6–37
- Formats available
- Transformers, GGUF, GGUF (imatrix)
- First indexed
- 13 Sept 2026
- Last updated
- 1 Nov 2026
Ablation technique
How MiniCPM5-2B-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: Layers 6–37. This comes from the publisher’s own model card and has not been independently verified.
From the publisher’s model card
“This is an uncensored version of openbmb/MiniCPM5-2B 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.
llama.cpp / GGUF
hf download tinyopsec/Huihui-MiniCPM5-2B-abliterated-GGUF --local-dir ./minicpm5-2b-abliterated
llama-cli -m ./minicpm5-2b-abliterated/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "huihui-ai/Huihui-MiniCPM5-2B-abliterated"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve huihui-ai/Huihui-MiniCPM5-2B-abliterated --trust-remote-codeDownloads and variants (9)
Every published repository of MiniCPM5-2B-abliterated, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| tinyopsec/Huihui-MiniCPM5-2B-abliterated-GGUF | GGUF | 1.7K |
| mradermacher/Huihui-MiniCPM5-2B-abliterated-i1-GGUF | GGUF (imatrix) | 909 |
| huihui-ai/Huihui-MiniCPM5-2B-abliteratedsource | Transformers | 774 |
| mradermacher/Huihui-MiniCPM5-2B-abliterated-GGUF | GGUF | 621 |
| edgebites/MiniCPM5-2B-Abliterated-iq4XS-GGUF | GGUF | 614 |
| mradermacher/MiniCPM5-2B-abliterated-i1-GGUF | GGUF (imatrix) | 354 |
| mradermacher/MiniCPM5-2B-abliterated-GGUF | GGUF | 278 |
| PinoCookie/MiniCPM5-2B-abliterated | Transformers | 97 |
| Vhh5/Huihui-MiniCPM5-2B-abliterated | Transformers | 68 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- MiniCPM5-2B-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 openbmb/MiniCPM5-2B, not added by ablation — at 2B parameters, expect a small model’s reasoning and a high error rate on exploit detail.
- Ablation is reported as limited to layers 6–37, so untouched layers may still reassert refusal on some phrasings.
- Declared licence is apache-2.0, but the terms that bind you are openbmb/MiniCPM5-2B’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 MiniCPM5-2B-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 huihui-ai, and does not independently verify publisher claims. See responsible use.