Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AutoCAP: Towards Automatic Cross-lingual Alignment Planning for Zero-shot Chain-of-Thought

Yongheng Zhang, Qiguang Chen, Min Li, Wanxiang Che, Libo Qin

TL;DR

An Automatic Cross-lingual Alignment Planning (AutoCAP) for zero-shot chain-of-thought achieves state-of-the-art performance, surpassing previous methods that required manual effort.

Abstract

Cross-lingual chain-of-thought can effectively complete reasoning tasks across languages, which gains increasing attention. Recently, dominant approaches in the literature improve cross-lingual alignment capabilities by integrating reasoning knowledge from different languages. Despite achieving excellent performance, current methods still have two main challenges: (1) Manual language specification: They still highly rely on manually selecting the languages to integrate, severely affecting their generalizability; (2) Static weight allocation: Current methods simply integrate all languages equally. In fact, different language reasoning paths should have different weights to achieve better complementation and integration. Motivated by this, we introduce an Automatic Cross-lingual Alignment Planning (AutoCAP) for zero-shot chain-of-thought to address the above challenges. The core of AutoCAP consists of two components: (1) Automatic Language Selection Prompting to guide LLMs to select appropriate languages and (2) Automatic Weight Allocation Prompting to automatically allocate alignment weight scores to each reasoning path. Extensive experiments on several benchmarks reveal that AutoCAP achieves state-of-the-art performance, surpassing previous methods that required manual effort.

AutoCAP: Towards Automatic Cross-lingual Alignment Planning for Zero-shot Chain-of-Thought

TL;DR

An Automatic Cross-lingual Alignment Planning (AutoCAP) for zero-shot chain-of-thought achieves state-of-the-art performance, surpassing previous methods that required manual effort.

Abstract

Cross-lingual chain-of-thought can effectively complete reasoning tasks across languages, which gains increasing attention. Recently, dominant approaches in the literature improve cross-lingual alignment capabilities by integrating reasoning knowledge from different languages. Despite achieving excellent performance, current methods still have two main challenges: (1) Manual language specification: They still highly rely on manually selecting the languages to integrate, severely affecting their generalizability; (2) Static weight allocation: Current methods simply integrate all languages equally. In fact, different language reasoning paths should have different weights to achieve better complementation and integration. Motivated by this, we introduce an Automatic Cross-lingual Alignment Planning (AutoCAP) for zero-shot chain-of-thought to address the above challenges. The core of AutoCAP consists of two components: (1) Automatic Language Selection Prompting to guide LLMs to select appropriate languages and (2) Automatic Weight Allocation Prompting to automatically allocate alignment weight scores to each reasoning path. Extensive experiments on several benchmarks reveal that AutoCAP achieves state-of-the-art performance, surpassing previous methods that required manual effort.
Paper Structure (26 sections, 8 equations, 8 figures, 2 tables)

This paper contains 26 sections, 8 equations, 8 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. Mono-lingual Chain-of-Thought
  4. Cross-lingual Chain-of-Thought
  5. Automatic Cross-lingual Alignment Planning
  6. Automatic Language Selection Prompting
  7. Automatic Weight Allocation Prompting
  8. Automatic Cross-lingual Prompting Consistency
  9. Experiments
  10. Dataset and Baseline
  11. Main Results
  12. Analysis
  13. Answer 1: All Modules in AutoCAP are Effective for AutoCAP
  14. Automatic Language Selection Prompting is effective.
  15. Automatic Weight Allocation Prompting is also crucial for performance enhancement.
  16. ...and 11 more sections

Figures (8)

  • Figure 1: Traditional Cross-lingual Self-consistent framework (a) vs. Automatic Cross-lingual Alignment Planner framework (b). Previous approaches require manually specifying the aligned languages (English, Spanish, and German), and assigning the same weights to these languages. In contrast, our framework (AutoCAP) uses the Automatic Language Selection and Automatic Weight Allocation to automatically select the most appropriate languages and weights.
  • Figure 2: The overall workflow of AutoCAP, which consist of Automatic Language Selection Prompting and Automatic Weight Allocation Prompting.
  • Figure 3: The accuracy of single-round AutoCAP compared to multi-round AutoCAP. AVG stands for average performance score.
  • Figure 4: Performance comparison results of CLSP qin2023cross and AutoCAP on different languages.
  • Figure 5: Languages for cross-lingual reasoning and their proportions.
  • ...and 3 more figures