NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-Uncensored
120BNemotronuncensored
from nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16
↓ 9.9K♥ 7
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara is a Heretic-ablated variant of nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16, published on Hugging Face by spiritfather. It is 120B total parameters (mixture-of-experts), server-class deployment class and other licence. It is distributed across 4 repositories in Transformers, GGUF (imatrix) and GGUF formats, totalling 9.9K downloads.
Specification
- Model name
- NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara
- Source repository
- spiritfather/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara
- Publisher
- spiritfather
- Parameters
- 120B (mixture-of-experts)
- VRAM required
- ~73.5 GB at Q4_K_M, 4k context (estimated) — runs on 80 GB cards
- Hardware class
- 100B+ — server-class deployment
- Licence
- other
- Task
- Image text to text
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- Transformers, GGUF (imatrix), GGUF
- First indexed
- 18 Oct 2026
- Last updated
- 19 Oct 2026
Ablation technique
How NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara 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
“🧬 Base model: nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download spiritfather/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara-i1-GGUF --local-dir ./nvidia-nemotron-3.5-super-vl-120b-a12b-heretic-ara
llama-cli -m ./nvidia-nemotron-3.5-super-vl-120b-a12b-heretic-ara/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "spiritfather/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve spiritfather/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara --trust-remote-codeDownloads and variants (3)
Every published repository of NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| spiritfather/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara-i1-GGUF | GGUF (imatrix) | 5.7K |
| spiritfather/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara-GGUF | GGUF | 4.1K |
| spiritfather/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-arasource | Transformers | 138 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara 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 nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16, not added by ablation — at 120B 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 other, but the terms that bind you are nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16’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 NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-heretic-ara is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with spiritfather, and does not independently verify publisher claims. See responsible use.