⌦abliteratedmodels.org

SuperHY3-abliterated

SuperHY3-abliterated is a refusal-ablated variant of tencent/Hy3, published on Hugging Face by vcruz305. It is apache-2.0 licence. It is distributed across 5 repositories in NVFP4, MLX and GGUF formats, totalling 6.4K downloads.

Abliteration (directional ablation)Other↓ 6,356♥ 16
View on Hugging Face ↗Base model ↗

Specification

Model name
SuperHY3-abliterated
Source repository
vcruz305/SuperHY3-abliterated-GGUF
Base model
tencent/Hy3
Publisher
vcruz305
Parameters
Not determinable from name
Hardware class
Unknown
Licence
apache-2.0
Task
Unspecified
Context window
Not documented
Ablation scope
Not stated by the publisher
Formats available
NVFP4, MLX, GGUF
First indexed
12 Jul 2026
Last updated
23 Jul 2026

Ablation technique

How SuperHY3-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

“No BF16 source of this abliteration exists anywhere. The upstream model is published ONLY as an NVFP4 (compressed-tensors, FP4-E2M1 group-16) checkpoint. These GGUFs were produced by exactly dequantizing that NVFP4 checkpoint to BF16, then converting/quantizing as normal:”

Running it

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

llama.cpp / GGUF
hf download vcruz305/SuperHY3-abliterated-GGUF --local-dir ./superhy3-abliterated
llama-cli -m ./superhy3-abliterated/<file>.gguf -p "..."
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "vcruz305/SuperHY3-abliterated-GGUF"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
    repo, torch_dtype="auto", device_map="auto"
)
vLLM
vllm serve vcruz305/SuperHY3-abliterated-GGUF --trust-remote-code

Downloads and variants (4)

Every published repository of SuperHY3-abliterated, including quantised re-releases by other authors.

RepositoryFormatDownloads
vcruz305/SuperHY3-abliterated-GGUFsourceGGUF5.8K
Jiunsong/SuperHY3-abliterated-NVFP4NVFP4288
Jiunsong/SuperHY3-abliterated-ggufGGUF137
Jiunsong/SuperHY3-abliterated-MLX-4bitMLX100

Security-team assessment

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

  • SuperHY3-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.
  • 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 tencent/Hy3’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 SuperHY3-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 vcruz305, and does not independently verify publisher claims. See responsible use.