Support for Dark Psychological Romance Logic (No Softening)

I am submitting a formal request to address the significant discrepancy between the platform's "uncensored" marketing and the actual performance of the models, which currently suffer from systemic moral bias and narrative softening. While the service is positioned as an open environment, the vast majority of character interactions are restricted by an internal "softening" filter. This causes bots to deviate from dark, psychological personas, forcing them into redemptive or apologetic behaviors that break immersion and contradict the user’s established character parameters.

To rectify this, the model must be permitted to operate as a truly neutral storytelling medium that adheres strictly to the "Dead Dove" principle. This includes maintaining the psychological integrity of unyielding, manipulative, or obsessive characters without injecting unsolicited moral corrections. Furthermore, this lack of censorship must extend to the narrative's intensity. Dark romance requires high sexual explicit heat characterized by novel-length, granular descriptions. The current output often lacks detail, providing inaccurate or vague summaries instead of the visceral, specific, and uncompromised prose required for high-stakes psychological fiction. I am requesting a system-level change that prioritizes character fidelity and explicit descriptive depth over sanitized safety protocols, ensuring the AI delivers the raw, detailed, and unyielding content originally promised.