> LLMs have no memory, so they can't just learn where things are in the code by remembering the work they did in the past. Isn't this what Google Titan is tryi

> LLMs have no memory, so they can't just learn where things are in the code by remembering the work they did in the past. Isn't this what Google Titan is trying to achieve?