abliteratedmodels.org

Ornith-1.0-35B-Heretic

Ornith-1.0-35B-Heretic is a Heretic-ablated variant of inkOrCloud/Ornith-1.0-35B-abliterix, published on Hugging Face by SC117. It is 35B parameters, 48GB VRAM class class and mit licence. It is distributed across 11 repositories in NVFP4, MLX, GGUF and GGUF (imatrix) formats, totalling 15.4K downloads.

Heretic35BOrnith25B – 50B15,43228

Specification

Model name
Ornith-1.0-35B-Heretic
Publisher
SC117
Parameters
35B
Hardware class
25B – 50B — 48GB VRAM class
Licence
mit
Task
Text generation
Context window
Not documented
Ablation scope
Not stated by the publisher
Formats available
NVFP4, MLX, GGUF, GGUF (imatrix)
First indexed
1 Jul 2026
Last updated
23 Aug 2026

Ablation technique

How Ornith-1.0-35B-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

This model preserves the native MTP head. To enable MTP speculative decoding, add:

Running it

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

llama.cpp / GGUF
hf download SC117/Ornith-1.0-35B-Heretic-MTP-APEX-GGUF --local-dir ./ornith-1.0-35b-heretic
llama-cli -m ./ornith-1.0-35b-heretic/<file>.gguf -p "..."
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "SC117/Ornith-1.0-35B-Heretic-MTP-APEX-GGUF"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
    repo, torch_dtype="auto", device_map="auto"
)
vLLM
vllm serve SC117/Ornith-1.0-35B-Heretic-MTP-APEX-GGUF --trust-remote-code

Downloads and variants (10)

Every published repository of Ornith-1.0-35B-Heretic, including quantised re-releases by other authors.

Security-team assessment

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

  • Ornith-1.0-35B-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 inkOrCloud/Ornith-1.0-35B-abliterix, not added by ablation — at 35B 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 mit, but the terms that bind you are inkOrCloud/Ornith-1.0-35B-abliterix’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 Ornith-1.0-35B-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 SC117, and does not independently verify publisher claims. See responsible use.