Selective Latent Thinking: Adaptive Compression of LLM Reasoning Chains

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xie, Hui, Liu, Jie, Qiao, Ziyue, Vanschore, Joaquin
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910254970699776
author Xie, Hui
Liu, Jie
Qiao, Ziyue
Vanschore, Joaquin
author_facet Xie, Hui
Liu, Jie
Qiao, Ziyue
Vanschore, Joaquin
contents Explicit chain-of-thought (CoT) reasoning substantially improves the reasoning ability of large language models (LLMs), but incurs high inference cost due to lengthy autoregressive traces. Existing latent reasoning methods offer a promising alternative, yet they often treat reasoning as uniformly compressible, causing precision-critical intermediate steps to be overly compressed and thereby degrading reasoning accuracy. In this work, we propose Selective Latent Thinking (SLT), a framework that selectively compresses redundant reasoning spans into latent representations while preserving precision-critical spans as explicit CoT within the same reasoning trajectory. Specifically, SLT first uses a lightweight decoder to anticipate a short upcoming reasoning span, and then applies confidence-based gating to determine the longest span that can be reliably compressed. The accepted span is encoded into a compact latent representation to improve reasoning efficiency, while uncertain or precision-critical reasoning remains in explicit CoT form to preserve accuracy. To learn this selective compression policy, SLT adopts a three-stage training strategy that combines span-level latent compression, reliability-aware future reasoning prediction, and trajectory-level reinforcement learning to optimize the trade-off between answer correctness and reasoning cost. Extensive experiments across four mathematical reasoning benchmarks demonstrate that SLT achieves 22.7\% higher accuracy than latent reasoning baselines at comparable compression ratios, while reducing reasoning chain length by 58.4\% with only 2.8\% accuracy degradation compared to explicit CoT,Our code can be found in https://github.com/hunshi34/SLT.
format Preprint
id arxiv_https___arxiv_org_abs_2605_25745
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Selective Latent Thinking: Adaptive Compression of LLM Reasoning Chains
Xie, Hui
Liu, Jie
Qiao, Ziyue
Vanschore, Joaquin
Computation and Language
Explicit chain-of-thought (CoT) reasoning substantially improves the reasoning ability of large language models (LLMs), but incurs high inference cost due to lengthy autoregressive traces. Existing latent reasoning methods offer a promising alternative, yet they often treat reasoning as uniformly compressible, causing precision-critical intermediate steps to be overly compressed and thereby degrading reasoning accuracy. In this work, we propose Selective Latent Thinking (SLT), a framework that selectively compresses redundant reasoning spans into latent representations while preserving precision-critical spans as explicit CoT within the same reasoning trajectory. Specifically, SLT first uses a lightweight decoder to anticipate a short upcoming reasoning span, and then applies confidence-based gating to determine the longest span that can be reliably compressed. The accepted span is encoded into a compact latent representation to improve reasoning efficiency, while uncertain or precision-critical reasoning remains in explicit CoT form to preserve accuracy. To learn this selective compression policy, SLT adopts a three-stage training strategy that combines span-level latent compression, reliability-aware future reasoning prediction, and trajectory-level reinforcement learning to optimize the trade-off between answer correctness and reasoning cost. Extensive experiments across four mathematical reasoning benchmarks demonstrate that SLT achieves 22.7\% higher accuracy than latent reasoning baselines at comparable compression ratios, while reducing reasoning chain length by 58.4\% with only 2.8\% accuracy degradation compared to explicit CoT,Our code can be found in https://github.com/hunshi34/SLT.
title Selective Latent Thinking: Adaptive Compression of LLM Reasoning Chains
topic Computation and Language
url https://arxiv.org/abs/2605.25745