Understand this story
On November 7, 2026, an uncensored fine-tuned model based on a distillation of DeepSeek-V4-Pro mixed with Qwen3.5 9B appeared in the open-source community. According to a report by A Yi AI Notes, the model refused only 3 of 100 red-team prompt tests; it uses Heretic's ARA algorithm to strip safety intervention vectors and was aligned with approximately 250,000 chain-of-thought reasoning data entries, with math and coding capabilities largely maintained. The model is packaged in Q4_K_M GGUF format, is 5.6GB in size, and can run locally on thin-and-light laptops with integrated graphics. There is currently no public evidence of further development progress.
AI-generated from public reports
Latest development · Report
Open-source community releases an uncensored fine-tuned model based on DeepSeek-V4-Pro and Qwen3.5 9BDevelopments
1 confirmed developments
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 7
AYi AI Notes
Open-source community releases an uncensored fine-tuned model based on DeepSeek-V4-Pro and Qwen3.5 9B
An uncensored fine-tuned model based on a hybrid of DeepSeek-V4-Pro distillation and Qwen3.5 9B has appeared in the open-source community, refusing only 3 out of 100 red-team prompt tests. The author says it uses Heretic's ARA algorithm to strip out safety intervention vectors and aligns with about 250,000 chain-of-thought reasoning data entries, with math and coding capabilities largely maintained; the model is packaged in Q4_K_M GGUF format at 5.6GB and can run locally on a thin-and-light laptop with integrated graphics.