Universal Transformers (2018)
A paper by Dehghani and colleagues proposed reusing transformer layers across depth, letting the model apply more computation per token. The design did not become a standard architecture.
The idea stayed a research curiosity.
It established the intuition that repeating layers can buy reasoning, which later looped models build on.
Inkling 2 Mini adds a learned gate that starts by blocking the looped message, then increases its weight during training.
