MiniCPM-V-4.7-1B-abliterated-heretic
1BMiniCPMheretic
from openbmb/MiniCPM-V-4.7-1B
↓ 12.1K♥ 14
MiniCPM-V-4.7-1B-heretic
MiniCPM-V-4.7-1B-heretic is a Heretic-ablated variant of openbmb/MiniCPM-V-4.7-1B, published on Hugging Face by saidutta69. It is 1B parameters, runs on a laptop CPU class and apache-2.0 licence. It is distributed across 5 repositories in GGUF and GGUF (imatrix) formats, totalling 17.7K downloads.
Specification
- Model name
- MiniCPM-V-4.7-1B-heretic
- Source repository
- saidutta69/MiniCPM-V-4.7-1B-heretic
- Base model
- openbmb/MiniCPM-V-4.7-1B
- Publisher
- saidutta69
- Parameters
- 1B
- VRAM required
- ~1.3 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
- Image text to text
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, GGUF (imatrix)
- First indexed
- 15 Oct 2026
- Last updated
- 1 Nov 2026
Ablation technique
How MiniCPM-V-4.7-1B-heretic was modified, and what that implies.
Heretic
Automated directional ablation via the Heretic toolchain, which searches for the refusal direction and applies it with a KL-divergence budget so general capability is preserved.
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
“A decensored variant of openbmb/MiniCPM-V-4.7-1B, produced with Heretic v1.4.0 (directional ablation / "abliteration"). Refusal behaviour is suppressed here via targeted weight edits to the attention output and MLP down-projections rather than fine-tuning, so the vision tower, the OCR and grounding behaviour, and the...”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download saidutta69/MiniCPM-V-4.7-1B-heretic --local-dir ./minicpm-v-4.7-1b-heretic
llama-cli -m ./minicpm-v-4.7-1b-heretic/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "saidutta69/MiniCPM-V-4.7-1B-heretic"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve saidutta69/MiniCPM-V-4.7-1B-heretic --trust-remote-codeDownloads and variants (4)
Every published repository of MiniCPM-V-4.7-1B-heretic, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| saidutta69/MiniCPM-V-4.7-1B-hereticsource | GGUF | 12K |
| mradermacher/MiniCPM-V-4.7-1B-heretic-i1-GGUF | GGUF (imatrix) | 3.6K |
| mradermacher/MiniCPM-V-4.7-1B-heretic-GGUF | GGUF | 1.3K |
| cecloud33/MiniCPM-V-4.7-1B-heretic-F16.gguf | GGUF | 799 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- MiniCPM-V-4.7-1B-heretic 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/MiniCPM-V-4.7-1B, not added by ablation — at 1B 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 apache-2.0, but the terms that bind you are openbmb/MiniCPM-V-4.7-1B’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 MiniCPM-V-4.7-1B-heretic is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with saidutta69, and does not independently verify publisher claims. See responsible use.