Esto es igual que el teorema de aproximacion universal. En la practica no dice nada de lo que puede lograrse. Es como decir que tienen un generador de numeros aleatorios.
What is the performance limit when scaling LLM inference? Sky's the limit.
We have mathematically proven that transformers can solve any problem, provided they are allowed to generate as many intermediate reasoning tokens as needed. Remarkably, constant depth is sufficient.
https://t.co/HO2seV73KT (ICLR 2024)