Feedback on Memory Summary
I shared some preferences about how I would like the AI to handle responses and memory. However, those preferences were mistakenly interpreted as part of my personal profile, and the following sentence was added to my Memory Summary:
“They also believe in analyzing the current situation and changing only what is necessary when evaluating information.”
This sentence does not represent my personality, values, or beliefs. It is actually a summary of my expectations for how I would like the AI to handle responses and memory, but it was incorrectly categorized as part of my personal profile.
Currently, I can add supplementary text to the Memory Summary, but I cannot remove, edit, replace, or regenerate an incorrect summary sentence. As a result, the only workaround is to add explanatory notes within my personal profile, which makes the summary feel unnatural and confusing.
What I want to edit is not my personal profile itself, but the specific sentence that the AI summarized incorrectly. I would appreciate a way to remove, replace, or regenerate only the incorrect summary, while leaving the rest of the profile unchanged.
I also believe that a user’s personal profile and their preferences for how the AI should behave are two different types of information. These should not be merged together during summarization.
Finally, I believe this issue could affect not only how the Memory Summary is displayed, but also how the AI understands and responds to the user in future conversations. Separating these two types of information would likely improve both the accuracy of the summary and the quality of future interactions.