Saved in:
Bibliographic Details
Main Authors: Xu, Yige, Guo, Xu, Zeng, Zhiwei, Miao, Chunyan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.12134
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909624405327872
author Xu, Yige
Guo, Xu
Zeng, Zhiwei
Miao, Chunyan
author_facet Xu, Yige
Guo, Xu
Zeng, Zhiwei
Miao, Chunyan
contents Chain-of-Thought (CoT) reasoning enables Large Language Models (LLMs) to solve complex reasoning tasks by generating intermediate reasoning steps. However, most existing approaches focus on hard token decoding, which constrains reasoning within the discrete vocabulary space and may not always be optimal. While recent efforts explore continuous-space reasoning, they often require full-model fine-tuning and suffer from catastrophic forgetting, limiting their applicability to state-of-the-art LLMs that already perform well in zero-shot settings with a proper instruction. To address this challenge, we propose a novel approach for continuous-space reasoning that does not require modifying the LLM. Specifically, we employ a lightweight fixed assistant model to speculatively generate instance-specific soft thought tokens as the initial chain of thoughts, which are then mapped into the LLM's representation space via a trainable projection module. Experimental results on five reasoning benchmarks demonstrate that our method enhances LLM reasoning performance through supervised, parameter-efficient fine-tuning. Source code is available at https://github.com/xuyige/SoftCoT.
format Preprint
id arxiv_https___arxiv_org_abs_2502_12134
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SoftCoT: Soft Chain-of-Thought for Efficient Reasoning with LLMs
Xu, Yige
Guo, Xu
Zeng, Zhiwei
Miao, Chunyan
Computation and Language
Chain-of-Thought (CoT) reasoning enables Large Language Models (LLMs) to solve complex reasoning tasks by generating intermediate reasoning steps. However, most existing approaches focus on hard token decoding, which constrains reasoning within the discrete vocabulary space and may not always be optimal. While recent efforts explore continuous-space reasoning, they often require full-model fine-tuning and suffer from catastrophic forgetting, limiting their applicability to state-of-the-art LLMs that already perform well in zero-shot settings with a proper instruction. To address this challenge, we propose a novel approach for continuous-space reasoning that does not require modifying the LLM. Specifically, we employ a lightweight fixed assistant model to speculatively generate instance-specific soft thought tokens as the initial chain of thoughts, which are then mapped into the LLM's representation space via a trainable projection module. Experimental results on five reasoning benchmarks demonstrate that our method enhances LLM reasoning performance through supervised, parameter-efficient fine-tuning. Source code is available at https://github.com/xuyige/SoftCoT.
title SoftCoT: Soft Chain-of-Thought for Efficient Reasoning with LLMs
topic Computation and Language
url https://arxiv.org/abs/2502.12134