One Model, Two Roles: Emergent Specialization in a Shared Recurrent Transformer
arXiv:2605.17811v2 Announce Type: replace-cross Abstract: Can a shared-weight recurrent Transformer---one that repeatedly applies the same Transformer to update evolving latent states---develop distinct internal roles without being partitioned into separate modules? We study this in Asymmetric Input Recurrence (AIR), a minimal two-state reasoning architecture in which the same Transformer model is reused for both updates (per literature, L and H) and the only built-in difference in the update r…