Decensored language models made using Heretic (https://github.com/p-e-w/heretic)
Philipp Emanuel Weidmann
p-e-w
AI & ML interests
Ethics of AI
Recent Activity
commented on
an
article
about 7 hours ago
Norm-Preserving Biprojected Abliteration
commented on
an
article
about 7 hours ago
Uncensor any LLM with abliteration
commented on
an
article
29 days ago
Uncensor any LLM with abliteration
Organizations
None yet
models
8
p-e-w/gemma-3-12b-it-heretic-v2
Image-Text-to-Text
•
12B
•
Updated
•
107
•
4
p-e-w/gpt-oss-20b-heretic-v2
Text Generation
•
21B
•
Updated
•
328
•
9
p-e-w/gemma-3-270m-it-heretic
Text Generation
•
0.3B
•
Updated
•
179
•
6
p-e-w/gpt-oss-20b-heretic
Text Generation
•
21B
•
Updated
•
934
•
71
p-e-w/Llama-3.1-8B-Instruct-heretic
Text Generation
•
8B
•
Updated
•
390
•
9
p-e-w/phi-4-heretic
Text Generation
•
15B
•
Updated
•
66
•
4
p-e-w/Qwen3-4B-Instruct-2507-heretic
Text Generation
•
4B
•
Updated
•
1.58k
•
25
p-e-w/gemma-3-12b-it-heretic
Image-Text-to-Text
•
12B
•
Updated
•
459
•
25
datasets
0
None public yet