SOURCE-LINKED INTELLIGENCE

RecurTrace: Adaptive Latent Reasoning with Loop-Time Memory

arXiv · AI, language, vision and robotics · article · Sep 3, 2026 · UTC

Repeating a small block of middle layers increases a language model's effective inference depth without adding parameters or generating extra tokens, and recent work shows that this latent recurrence improves reasoning. However, two design choices limit these gains. Each iteration sees only the previous output and cannot directly access earlier computations. Moreover, a fixed loop count wastes depth on easy inputs while leaving hard ones with too little computation. We introduce RecurTrace, which addresses both limitations using the loop's own trajectory. Specifically, Loop Memory Attention le

Read original source ↗ Open in workspace

recordType

paper

region

Global