Mohammad Saffar, a research scientist at Google DeepMind, shared a post on X highlighting prior research on universal transformers. He stated that looping transformers should not be considered new and pointed to work by m__dehghani and coauthors. Saffar indicated this paper appeared shortly after the initial transformer publication. The message serves as a reminder of earlier explorations into transformer variants with iterative components. Visible discussion in the post centers on crediting foundational papers in multimodal and architecture research. The claim remains the author's attribution rather than a verified new finding.