Multiple Regenerations Causing Context Reversion

Hi Venice Team,

I am writing to report a recurring issue I am experiencing with the web version of Venice that significantly impacts the user experience for roleplay and exploration.

The Issue:
I was using the Venice model for a text-based roleplay session to explore a world. However, I encountered some logical inconsistencies in the AI's response. To fix this, I clicked the "Regenerate" button.

The problem occurs immediately after regeneration: Both the old (unsatisfactory) answer and the newly generated answer are visible in the chat interface. However, when I continue the conversation by sending a new prompt, the AI's context completely ignores the new version and reverts back to using the old answer that I was dissatisfied with.

Expected Behavior:
The AI's context should lock onto the latest generated version so that the conversation can proceed logically from the improved response, not the one I tried to replace.

This issue makes the chat experience very frustrating, as it feels like the regeneration feature is not actually updating the conversation state. I hope you can look into this and fix it soon!

Platform: Venice Web (Browser)