Blog
Notes on research I have done and ideas I keep coming back to.
Explore looped Transformers through Classification by Attention (2020): shared weights, schema feedback, and a precise comparison with Meta’s looped LLM research.