Hello,
I’ve encountered an issue related to the 400 “Repetitive Patterns” error. The problem arises when a user sends messages that include previous conversation history. In this scenario, it’s unclear which specific message or sender is responsible for triggering the error.
For example, if a user requests the model to generate a message with the word “hello” repeated 200 times, the AI’s response would be included in the conversation history. When the next message is sent, the model gets triggered by its own previous response, effectively causing the user to softlock themselves.
This can also happen with the content filter error, especially when switching between models where the content filter is either enabled or disabled. In both cases, the user is stuck without a clear way to identify which part of the message history is causing the issue.
Are there any guidelines or best practices for resolving this situation without manually going through and testing each message? It would be helpful to know how to handle these errors more efficiently.
Thank you in advance for your help!