Abliteration Is Not a Scalpel: Off-Target Effects of Refusal Removal on Decision Disposition Across Model Families
Paper • 2607.17427 • Published
Aleksander Fafuła
oleczek
AI & ML interests
Behavioral psychometrics of LLMs - measuring what weight edits (abliteration, fine-tuning) do to a model's disposition. Multi-agent LLM systems on real financial markets.
Recent Activity
authored a paper 12 days ago
Abliteration Is Not a Scalpel: Off-Target Effects of Refusal Removal on Decision Disposition Across Model Families new activity 12 days ago
huihui-ai/Huihui-Qwen3-30B-A3B-Instruct-2507-abliterated:Update to README.md new activity 12 days ago
huihui-ai/Huihui-gemma-4-26B-A4B-it-abliterated:Update README.mdOrganizations
None yet
authored a
paper 12 days ago
Update to README.md
#4 opened 12 days ago
by oleczek
Update README.md
#4 opened 12 days ago
by oleczek