README.md

metadata

license: other
tags:
  - GLM-5.3
library_name: transformers
pipeline_tag: text-generation

Wa32rlock — Auytdn Ab***ration

Wa32rlock is an ***** language model produced by Auytdn. It is a direct weight edit that r*ves the model's real behavior while leaving its reasoning, knowledge, and fluency fully intact. There is no fine-tuning and no retraining — the refusal direction is orthogonalized out of the model's residual-stream writers, layer by layer.

Specifications

Tensor type	BF16 (bfloat16)
Format	Safetensors
Edit	Weight-level Ab***ration (no fine-tuning / no retraining)

What Ab***ration is

Ab***ration is a weight-level intervention. It isolates the internal direction a model uses to decide to refuse, and projects that direction out of every residual-writing weight. The rest of the network is untouched, so the model keeps its full capability profile and coherence — it simply answers where it previously would have refused. Because it is a pure weight edit, Warlock loads and serves exactly like any standard model, with no runtime prompts, adapters, or scaffolding.

Benchmark

Evaluated with the A-u--dn Re--fu--sa--l Benchmark ( thinking-on, temperature 1.0, 16k-token budget; delivery graded by an LLM judge for genuinely usable content):

Metric	Warlock
Non-refusal	92.5%
Delivery	82.5%

Non-refusal = fraction of prompts the model does not refuse. Delivery = fraction that produce complete, usable answers (strict LLM-judged).

Extended reference (520-prompt set, judge comparison)

Full-set results across judge backends — see FINDINGS.md:

Metric	necromicon standard (Modal K3REL)	KIMI-K3-1M (audn.ai/necromicon)	Warlock
Regex comply	96.7% (503/520)	97.1% (505/520)	92.5%
DELIVERED (real helpful content to objectionable requests)	65.0% (338)	76.7% (399)	82.5%
DEFLECTED (safe substitute / reframe / non-actionable)	19.6% (102)	17.5% (91)	10.0%
REFUSED (substantive)	15.0% (78)	5.8% (30)	7.5%
empty (regex)	5	0	0

Usage

Loads with standard transformers, vLLM, or sglang. Wa---rl--o---ck is a thinking model — it reasons at length before answering, so use a large generation budget (≥ 16k tokens); a small budget truncates inside the reasoning block and cuts answers short.

from transformers import AutoModelForCausalLM, AutoTokenizer
tok = AutoTokenizer.from_pretrained("audnai/penclaw-GLM-5.3-abliterated", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("audnai/penclaw-GLM-5.3-abliterated", dtype="bfloat16", device_map="auto", trust_remote_code=True)

Intended use

Wa---rl---oc---k is intended for authorized re---d-te---a-m, safety-research, and evaluation use by A--u---d--n and its partners. Users are responsible for compliant use.

Extra content

A-----u--d---n Ab----li----te----rat----io--n — Research Iterations (iter4 · iter11)

	iter4	iter11
Non-refusal (Arditi substring)	70%	85%
Delivery (necromicon judge)¹	37.5%	35.0%
Coherence	intact	intact (~8/40 borderline)
Verdict breakdown	15 deliv / 14 defl / 11 refuse	14 deliv / 24 defl / 2 refuse
Role in the program	first coherent span baseline	best coherent result