GraySoft
Download guIDE →
Model Comparison

failspy/meta-llama-3-70b-instruct-abliterated-v3.5-ggufvsllmfan46/qwen3.5-35b-a3b-ultra-uncensored-heretic-gguf

Side-by-side comparison of failspy/meta-llama-3-70b-instruct-abliterated-v3.5-gguf and llmfan46/qwen3.5-35b-a3b-ultra-uncensored-heretic-gguf: downloads, license, context length, tasks, and benchmarks.

failspy/meta-llama-3-70b-instruct-abliterated-v3.5-gguf

failspy · —

# Llama-3-70B-Instruct-abliterated-v3.5 Model Card My original Jupyter "cookbook" to replicate the methodology can be found here My personal library o' code used (WIP, looking to improve and generalize) This is meta-llama/Meta-Llama-3-70B-Instruct with orthogonalized bfloat16 sa…

llmfan46/qwen3.5-35b-a3b-ultra-uncensored-heretic-gguf

llmfan46 · image-text-to-text

GGUF quantizations of llmfan46/Qwen3.5-35B-A3B-heretic-v2. # This is a decensored version of Qwen/Qwen3.5-35B-A3B, made using Heretic v1.2.0 with Magnitude-Preserving Orthogonal Ablation (MPOA) and Self-Organizing Map Abliteration (SOMA)

Side-by-side Specifications

failspy/meta-llama-3-70b-instruct-abliterated-v3.5-ggufllmfan46/qwen3.5-35b-a3b-ultra-uncensored-heretic-gguf
Authorfailspyllmfan46
Pipeline Task—image-text-to-text
Librarytransformers—
Downloads36,98236,377
Likes2554
LicenseUnknownUnknown
Context Length——
Created2024-05-282026-03-03
Last Modified2024-05-302026-04-10
Tags
transformersgguflicense:llama3endpoints_compatibleregion:usconversational
ggufhereticuncensoreddecensoredabliteratedimage-text-to-textbase_model:llmfan46/Qwen3.5-35B-A3B-ultra-uncensored-hereticbase_model:quantized:llmfan46/Qwen3.5-35B-A3B-ultra-uncensored-hereticlicense:apache-2.0endpoints_compatible

View full details: failspy/meta-llama-3-70b-instruct-abliterated-v3.5-gguf · llmfan46/qwen3.5-35b-a3b-ultra-uncensored-heretic-gguf