base_model:

• llmfan46/Forgotten-Transgression-24B-v4.1-uncensored-heretic base_model_relation: quantized language:
• en license: apache-2.0 inference: false tags:
• nsfw
• explicit
• roleplay
• unaligned
• dangerous
• ERP
• heretic
• uncensored
• decensored
• abliterated
• ara

94% fewer refusals (6/100 Uncensored vs 95/100 Original) while preserving model quality (0.0232 KL divergence).

❤️ Support My Work

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

image/png

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.

GGUF quantizations of llmfan46/Forgotten-Transgression-24B-v4.1-uncensored-heretic.

This is a decensored version of ReadyArt/Forgotten-Transgression-24B-v4.1, made using Heretic v1.2.0 with the Arbitrary-Rank Ablation (ARA) method

Abliteration parameters

Targeted components

• attn.o_proj

Performance

PIQA test results with batch size 128:

Original:

Heretic:

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 32:

Original:

Heretic:

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

Quantizations

Usage

Works with llama.cpp, LM Studio, Ollama, and other GGUF-compatible tools.