Recurrent Looped Transformer paper: decoder recurs across every token for unbounded latent depth
NandoDF · x · 2026-09-13
A new paper proposes Recurrent Looped Transformer (RLT), an architecture where the decoder runs recurrently across every prompt and response token, carrying its final hidden state and a sliding-window attention cache forward, while a causal encoder builds reusable global KV memory.
Key points:
- Unbounded latent depth: sequences grow computation paths without adding physical layers — a fixed number of blocks per token.
- Three principles: unbounded temporal depth for latent reasoning, model–hardware co-design (separating parallel encoder work from recurrent decoder work, enabling batching, memory reuse, checkpointed training), and model–RL algorithm co-design.
- RL consistency: pretraining, SFT, rollout sampling and current-policy replay share the same state transition, and exact replay reconstructs states under current parameters instead of reusing stale rollout states.
The author positions RLT as a concrete basis for hardware-aware recurrent execution and reliable RL scaling.
More from Infra
- How Different NVIDIA GPUs Change Model Outputs via BF16, FP8, FP4 Precision — Kyrannio · 2026-10-31
- Kubernetes AI Conformance becomes a permanent subproject under SIG Architecture — TerryTangYuan · 2026-10-31
- Google shares first AI-generated response from space — a haiku, per Pichai — LinkedInNews · 2026-10-31
- Debugging llama.cpp: sliding-window context checkpoints silently eat your RAM — jomi-se · 2026-10-31
- Cornell Professor's Startup Makora Wins $1.75M to Make AI Faster and Cheaper to Run — srush_nlp · 2026-10-31
- Transformers.js monthly npm downloads grow 9x in a year, from 1.64M to ~15M — nicodotdev · 2026-10-31