Why is it that for any model in chat mode, after a specific amount of messages the model starts spitting the same answer?

[ye7iaserag]

This has happened to me with every single model I've used, last model I tried is gpt4-x-alpaca-13b-native-4bit-128g and I think it's the best model that I've tried yet but it still suffers the same problem.
I've also tried with multiple presets and configs, yet I would always hit this wall

Why does that happen or what leads to it?
Let's discuss maybe we hit something that allows us to fix this for good.