AKME groups conversation in threads.
A thread is a conversation on a specific topic. Ask AKME a question and the GPT will answer.
If you have a follow-up question, continue in the same thread. AKME will also send the message history with your new question, so the GPT will have context (will "remember").
If instead you want to ask something different, go back and start a new thread. This avoids sending the message history - and spending tokens - when it's not relevant for the question.
Look how they ...
The answer from AKME stops mid-sentence (or mid-word)?
Simply ask AKME to continue
This happens because the answer reached the "max tokens" limit set in the Parameters.
If you encounter this issue repeatedly, you can adjust the parameters and increase the "Max Tokens" value.