Recurrent Looped Transformer feeds final hidden states back for deeper reasoning
WebAssemblyMan · reddit · 2026-09-16
Recurrent Looped Transformer (RLT) passes the decoder's final hidden state to the next token alongside its causal encoder representation, reads global KV memory, and maintains a sliding-window attention cache at every layer — the same update runs over prompt and response tokens. The claimed benefit is more effective reasoning depth; code is on GitHub.
More from Research
- Radial gate objective cuts diffusion transformer FID from 16.24 to 11.83 — Junha Park · 2026-11-02
- REA: async brain-cerebellum runtime for robots where policies never wait on the LLM — Pengfei Zhou · 2026-11-02
- Microsoft's MemSteer turns unlabeled user histories into agent memory supervision — MicrosoftResearch · 2026-11-02
- SciGame benchmark uses counterfactual worlds to expose LLM agents' priors vs evidence — Yusheng Zhao · 2026-11-02
- On Terence Tao's blog, Frank Calegari calls on mathematicians to adapt after OpenAI's 700-paper release — Hesamation · 2026-11-02
- Tsinghua & Microsoft's HarnessMM cuts agent token use by up to 87.93% while raising success rate — rohanpaul_ai · 2026-11-02