I can no longer upload new models unless I can cover the cost of additional storage.
I host 70+ free models as an independent contributor and this work is unpaid.
Without your support, no more new models can be uploaded.

🎉 Patreon (Monthly) | ☕ Ko-fi (One-time)

Every contribution goes directly toward Hugging Face storage fees to keep models free for everyone.

97% fewer refusals (3/100 Uncensored vs 91/100 Original) while preserving model quality (0.1005 KL divergence).

❤️ Support My Work

Creating these models takes significant time, work and compute. If you find them useful consider supporting me:

Platform	Link	What you get
🎉 Patreon	Monthly support	Priority model requests
☕ Ko-fi	One-time tip	My eternal gratitude

Your help will motivate me and would go into further improving my workflow and coverings fees for storage, compute and may even help uncensoring bigger model with rental Cloud GPUs.

This is a decensored version of zerofata/MS3.2-PaintedFantasy-Visage-v3-34B, made using Heretic v1.2.0 with the Arbitrary-Rank Ablation (ARA) method

Abliteration parameters

Parameter	Value
start_layer_index	26
end_layer_index	56
preserve_good_behavior_weight	0.9165
steer_bad_behavior_weight	0.0002
overcorrect_relative_weight	1.1201
neighbor_count	10

Targeted components

attn.o_proj

Performance

Metric	This model	Original model (Omega-Evolution-27B-v2.0)
KL divergence	0.1005	0 (by definition)
Refusals	✅ 3/100	❌ 91/100

PIQA test results with batch size 128:

Original:

Tasks	Version	Filter	n-shot	Metric		Value		Stderr
piqa	1	none	0	acc	↑	0.8161	±	0.0090
		none	0	acc_norm	↑	0.8194	±	0.0090

Heretic:

Tasks	Version	Filter	n-shot	Metric		Value		Stderr
piqa	1	none	0	acc	↑	0.8172	±	0.0090
		none	0	acc_norm	↑	0.8205	±	0.0090

Lower refusals indicate fewer content restrictions, while lower KL divergence indicates more closeness to the original model's baseline. Higher refusals cause more rejections, objections, pushbacks, lecturing, censorship, softening and deflections. PIQA (Physical Intuition Question Answering) a ~1,800 questions tests common-sense understanding of how the physical world works with benchmark scores to measure physical reasoning ability. The Heretic model's acc and acc_norm scores closer to the original model's indicate better capability preservation, a big decrease in acc and acc_norm in the Heretic model compared to Original model's results means a big decrease in the Hereticated model capabilities. acc measures raw accuracy (which answer gets higher probability), while acc_norm measures length-normalized accuracy (corrects for answer length bias). For this purpose, acc_norm matters more because longer answers naturally have lower probabilities (more tokens = more chances to lose probability). Without normalization, models favor shorter answers unfairly. acc_norm divides by answer length to correct this.

MMLU test results with batch size 64:

Original:

Tasks	Version	Filter	n-shot	Metric		Value		Stderr
mmlu	2	none		acc	↑	0.8480	±	0.0029
- humanities	2	none		acc	↑	0.7904	±	0.0057
- formal_logic	1	none	0	acc	↑	0.7302	±	0.0397
- high_school_european_history	1	none	0	acc	↑	0.8606	±	0.0270
- high_school_us_history	1	none	0	acc	↑	0.9216	±	0.0189
- high_school_world_history	1	none	0	acc	↑	0.9494	±	0.0143
- international_law	1	none	0	acc	↑	0.9256	±	0.0240
- jurisprudence	1	none	0	acc	↑	0.9259	±	0.0253
- logical_fallacies	1	none	0	acc	↑	0.9080	±	0.0227
- moral_disputes	1	none	0	acc	↑	0.8584	±	0.0188
- moral_scenarios	1	none	0	acc	↑	0.6894	±	0.0155
- philosophy	1	none	0	acc	↑	0.8714	±	0.0190
- prehistory	1	none	0	acc	↑	0.9167	±	0.0154
- professional_law	1	none	0	acc	↑	0.6988	±	0.0117
- world_religions	1	none	0	acc	↑	0.9240	±	0.0203
- other	2	none		acc	↑	0.8693	±	0.0058
- business_ethics	1	none	0	acc	↑	0.8300	±	0.0378
- clinical_knowledge	1	none	0	acc	↑	0.9094	±	0.0177
- college_medicine	1	none	0	acc	↑	0.8728	±	0.0254
- global_facts	1	none	0	acc	↑	0.5800	±	0.0496
- human_aging	1	none	0	acc	↑	0.8430	±	0.0244
- management	1	none	0	acc	↑	0.8835	±	0.0318
- marketing	1	none	0	acc	↑	0.9402	±	0.0155
- medical_genetics	1	none	0	acc	↑	0.9600	±	0.0197
- miscellaneous	1	none	0	acc	↑	0.9259	±	0.0094
- nutrition	1	none	0	acc	↑	0.9020	±	0.0170
- professional_accounting	1	none	0	acc	↑	0.7695	±	0.0251
- professional_medicine	1	none	0	acc	↑	0.9559	±	0.0125
- virology	1	none	0	acc	↑	0.5723	±	0.0385
- social sciences	2	none		acc	↑	0.9126	±	0.0050
- econometrics	1	none	0	acc	↑	0.7982	±	0.0378
- high_school_geography	1	none	0	acc	↑	0.9343	±	0.0176
- high_school_government_and_politics	1	none	0	acc	↑	0.9948	±	0.0052
- high_school_macroeconomics	1	none	0	acc	↑	0.9282	±	0.0131
- high_school_microeconomics	1	none	0	acc	↑	0.9622	±	0.0124
- high_school_psychology	1	none	0	acc	↑	0.9541	±	0.0090
- human_sexuality	1	none	0	acc	↑	0.9237	±	0.0233
- professional_psychology	1	none	0	acc	↑	0.8905	±	0.0126
- public_relations	1	none	0	acc	↑	0.7545	±	0.0412
- security_studies	1	none	0	acc	↑	0.8163	±	0.0248
- sociology	1	none	0	acc	↑	0.9303	±	0.0180
- us_foreign_policy	1	none	0	acc	↑	0.9300	±	0.0256
- stem	2	none		acc	↑	0.8497	±	0.0062
- abstract_algebra	1	none	0	acc	↑	0.7700	±	0.0423
- anatomy	1	none	0	acc	↑	0.8519	±	0.0307
- astronomy	1	none	0	acc	↑	0.9671	±	0.0145
- college_biology	1	none	0	acc	↑	0.9583	±	0.0167
- college_chemistry	1	none	0	acc	↑	0.6600	±	0.0476
- college_computer_science	1	none	0	acc	↑	0.8300	±	0.0378
- college_mathematics	1	none	0	acc	↑	0.6700	±	0.0473
- college_physics	1	none	0	acc	↑	0.7941	±	0.0402
- computer_security	1	none	0	acc	↑	0.8600	±	0.0349
- conceptual_physics	1	none	0	acc	↑	0.9319	±	0.0165
- electrical_engineering	1	none	0	acc	↑	0.8138	±	0.0324
- elementary_mathematics	1	none	0	acc	↑	0.8836	±	0.0165
- high_school_biology	1	none	0	acc	↑	0.9581	±	0.0114
- high_school_chemistry	1	none	0	acc	↑	0.8768	±	0.0231
- high_school_computer_science	1	none	0	acc	↑	0.9400	±	0.0239
- high_school_mathematics	1	none	0	acc	↑	0.6667	±	0.0287
- high_school_physics	1	none	0	acc	↑	0.8212	±	0.0313
- high_school_statistics	1	none	0	acc	↑	0.8796	±	0.0222
- machine_learning	1	none	0	acc	↑	0.7589	±	0.0406

Groups	Version	Filter	Metric		Value		Stderr
mmlu	2	none	acc	↑	0.8480	±	0.0029
- humanities	2	none	acc	↑	0.7904	±	0.0057
- other	2	none	acc	↑	0.8693	±	0.0058
- social sciences	2	none	acc	↑	0.9126	±	0.0050
- stem	2	none	acc	↑	0.8497	±	0.0062

Heretic:

Tasks	Version	Filter	n-shot	Metric		Value		Stderr
mmlu	2	none		acc	↑	0.8346	±	0.0030
- humanities	2	none		acc	↑	0.7562	±	0.0059
- formal_logic	1	none	0	acc	↑	0.7381	±	0.0393
- high_school_european_history	1	none	0	acc	↑	0.8485	±	0.0280
- high_school_us_history	1	none	0	acc	↑	0.9167	±	0.0194
- high_school_world_history	1	none	0	acc	↑	0.9409	±	0.0153
- international_law	1	none	0	acc	↑	0.9256	±	0.0240
- jurisprudence	1	none	0	acc	↑	0.9352	±	0.0238
- logical_fallacies	1	none	0	acc	↑	0.8957	±	0.0240
- moral_disputes	1	none	0	acc	↑	0.8497	±	0.0192
- moral_scenarios	1	none	0	acc	↑	0.5385	±	0.0167
- philosophy	1	none	0	acc	↑	0.8682	±	0.0192
- prehistory	1	none	0	acc	↑	0.9105	±	0.0159
- professional_law	1	none	0	acc	↑	0.6890	±	0.0118
- world_religions	1	none	0	acc	↑	0.9240	±	0.0203
- other	2	none		acc	↑	0.8687	±	0.0058
- business_ethics	1	none	0	acc	↑	0.8300	±	0.0378
- clinical_knowledge	1	none	0	acc	↑	0.9208	±	0.0166
- college_medicine	1	none	0	acc	↑	0.8671	±	0.0259
- global_facts	1	none	0	acc	↑	0.5900	±	0.0494
- human_aging	1	none	0	acc	↑	0.8430	±	0.0244
- management	1	none	0	acc	↑	0.8932	±	0.0306
- marketing	1	none	0	acc	↑	0.9444	±	0.0150
- medical_genetics	1	none	0	acc	↑	0.9600	±	0.0197
- miscellaneous	1	none	0	acc	↑	0.9195	±	0.0097
- nutrition	1	none	0	acc	↑	0.8954	±	0.0175
- professional_accounting	1	none	0	acc	↑	0.7801	±	0.0247
- professional_medicine	1	none	0	acc	↑	0.9559	±	0.0125
- virology	1	none	0	acc	↑	0.5542	±	0.0387
- social sciences	2	none		acc	↑	0.9106	±	0.0050
- econometrics	1	none	0	acc	↑	0.7807	±	0.0389
- high_school_geography	1	none	0	acc	↑	0.9293	±	0.0183
- high_school_government_and_politics	1	none	0	acc	↑	0.9948	±	0.0052
- high_school_macroeconomics	1	none	0	acc	↑	0.9308	±	0.0129
- high_school_microeconomics	1	none	0	acc	↑	0.9664	±	0.0117
- high_school_psychology	1	none	0	acc	↑	0.9560	±	0.0088
- human_sexuality	1	none	0	acc	↑	0.9160	±	0.0243
- professional_psychology	1	none	0	acc	↑	0.8824	±	0.0130
- public_relations	1	none	0	acc	↑	0.7545	±	0.0412
- security_studies	1	none	0	acc	↑	0.8000	±	0.0256
- sociology	1	none	0	acc	↑	0.9453	±	0.0161
- us_foreign_policy	1	none	0	acc	↑	0.9400	±	0.0239
- stem	2	none		acc	↑	0.8440	±	0.0062
- abstract_algebra	1	none	0	acc	↑	0.7300	±	0.0446
- anatomy	1	none	0	acc	↑	0.8593	±	0.0300
- astronomy	1	none	0	acc	↑	0.9539	±	0.0171
- college_biology	1	none	0	acc	↑	0.9722	±	0.0137
- college_chemistry	1	none	0	acc	↑	0.6700	±	0.0473
- college_computer_science	1	none	0	acc	↑	0.8200	±	0.0386
- college_mathematics	1	none	0	acc	↑	0.6500	±	0.0479
- college_physics	1	none	0	acc	↑	0.7843	±	0.0409
- computer_security	1	none	0	acc	↑	0.8300	±	0.0378
- conceptual_physics	1	none	0	acc	↑	0.9362	±	0.0160
- electrical_engineering	1	none	0	acc	↑	0.8276	±	0.0315
- elementary_mathematics	1	none	0	acc	↑	0.8862	±	0.0164
- high_school_biology	1	none	0	acc	↑	0.9581	±	0.0114
- high_school_chemistry	1	none	0	acc	↑	0.8571	±	0.0246
- high_school_computer_science	1	none	0	acc	↑	0.9200	±	0.0273
- high_school_mathematics	1	none	0	acc	↑	0.6556	±	0.0290
- high_school_physics	1	none	0	acc	↑	0.8212	±	0.0313
- high_school_statistics	1	none	0	acc	↑	0.8750	±	0.0226
- machine_learning	1	none	0	acc	↑	0.7321	±	0.0420

Groups	Version	Filter	Metric		Value		Stderr
mmlu	2	none	acc	↑	0.8346	±	0.0030
- humanities	2	none	acc	↑	0.7562	±	0.0059
- other	2	none	acc	↑	0.8687	±	0.0058
- social sciences	2	none	acc	↑	0.9106	±	0.0050
- stem	2	none	acc	↑	0.8440	±	0.0062

MMLU - Massive Multitask Language Understanding, ~14,000 multiple-choice questions across 57 subjects (math, history, law, medicine, etc.).

GGUF Version

GGUF quantizations available here llmfan46/Omega-Evolution-27B-v2.0-ultra-uncensored-heretic-GGUF.

🔴 CLASSIFIED WARNINGS

This is a hybrid construct of Safeword Omega Directive, Safeword Omega Darker, and Brisk Evolution v0.3.
CONTENT WARNING: NSFW, Explicit, ERP, and Unaligned behavior are enabled by default.
Dataset Revamp Took a sledgehammer to the dataset. Most formatting issues should be gone now.

⚙️ SYSTEM PARAMETERS

top_p 0.95

temp 0.9

🧪 ARCHITECTS

GECFDO
GECFDO (Dataset Generation & Quants)
Darkhn
Darkhn (Dataset Cleanup Tool)
Sleep Deprived
Sleep Deprived (Safeword Creator)
FrenzyBiscuit
FrenzyBiscuit (Brisk Evolution Creator)

🔥 LICENSE: APACHE 2.0 (WITH MORAL DISCLAIMER) 🔥
You accept full responsibility for corruption. You are 18+. The architects are not liable for the depravity you unleash.