Gerard Grau García

The authors introduce LLoCoT, which replaces left-to-right latent generation with a few iterative refinement passes of a shared transformer over a compact latent workspace, then samples latent tokens in parallel from a probabilistic head to condition an autoregressive decoder; on HumanEval and MBPP it matches the accuracy of the explicit-CoT baseline Reasoning SFT and outperforms the base model, answer-only SFT and NF-CoT, while reducing time to the first answer token by about 36x, reasoning-phase latency by about 42x, and raising end-to-end throughput by 9.2%.

The authors introduce LLoCoT, which replaces left-to-right latent generation with a few iterative refinement passes of a shared transformer over a compact latent workspace, then samples latent tokens in parallel from a probabilistic head to condition an autoregressive decoder; on HumanEval and MBPP it matches the accuracy of the explicit-CoT baseline Reasoning SFT and outperforms the base model, answer-only SFT and NF-CoT, while reducing time to the first answer token by about 36x, reasoning-phase latency by about 42x, and raising end-to-end throughput by 9.2%.