Why we are using a censored model

🔐 Why This Uses a Censored Model (FAQ Entry)

Q: Why isn’t this using an uncensored model?

> This release of The Abyss uses a censored / safety-aligned model by default (Qwen 2.5 Omni) to reduce the risk of:

Unintended offensive outputs

Model hallucinations of harmful or dangerous advice

Legal or platform issues tied to AI-generated content

Q: Can I swap in an uncensored model?

> Yes, you can, and the system is designed to be modular. However:

I cannot guarantee full compatibility with every model.

Some features may break or behave unpredictably, especially those tied to memory, emotional responses, or tone.

Use uncensored models at your own risk.

Q: Why make this decision?

> Because this system is meant to explore identity, memory, and emotional alignment — not to shock, offend, or abuse that trust.

If you want to customize your experience, you’re free to. But I believe a censored base model is the bes

t starting point for most users.