← Back to the wire

Recurrent Looped Transformer

AnnouncementModelSep 12, 2026

The Recurrent Looped Transformer (RLT) pairs a causal encoder with a recurrent decoder whose hidden state and sliding-window attention cache carry across all prompt and response tokens. In its concrete configuration, the model uses 48 encoder layers and 48 decoder layers, so the temporal path traverses 48t decoder blocks after t tokens. The paper notes that realized reasoning gains, hardware efficiency, and RL scaling remain to be established.

Receipt № 18821 source · awaiting confirmation ◐

Evidence

1source· awaiting independent confirmation

No score is assigned. Sources and their independence are shown in the citation chain below.

Citation chain · 1 source

Recurrent Looped TransformerModel
Canonical: https://yifanzhang-pro.github.io/recurrent-looped-tranformer/