Beyond Chain-of-Thought: A Survey of Chain-of-X Paradigms for LLMs

2024-04-24

International Conference on Computational Linguistics

AI Summary

This work addresses the fragmented development of Chain-of-X (CoX) methods for enhancing large language model (LLM) reasoning. We propose the first systematic, unified CoX taxonomy, organizing over 30 variants—including Chain-of-Thought (CoT), Chain-of-Symbol (CoS), and Chain-of-Uncertainty (CoU)—along two orthogonal dimensions: *node type* (X) and *task scenario*. The framework encompasses 12 X-node categories (e.g., evidence, action, uncertainty) and five major application domains. Through integrated analysis of prompt engineering, cognitive modeling, and empirical evaluation, we uncover the underlying modeling principles, applicability boundaries, and cross-task transfer patterns of distinct X-nodes, distilling reusable prompt design guidelines. We further establish a structured evaluation methodology to assess CoX efficacy rigorously. Our contributions provide both theoretical foundations and practical guidance for interpretable LLM reasoning and domain-adaptive inference.

Technology Category

• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Reasoning under Uncertainty: Causality

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Chain-of-Thought (CoT) has been a widely adopted prompting method, eliciting impressive reasoning abilities of Large Language Models (LLMs). Inspired by the sequential thought structure of CoT, a number of Chain-of-X (CoX) methods have been developed to address various challenges across diverse domains and tasks involving LLMs. In this paper, we provide a comprehensive survey of Chain-of-X methods for LLMs in different contexts. Specifically, we categorize them by taxonomies of nodes, i.e., the X in CoX, and application tasks. We also discuss the findings and implications of existing CoX methods, as well as potential future directions. Our survey aims to serve as a detailed and up-to-date resource for researchers seeking to apply the idea of CoT to broader scenarios.

Problem

• Survey Chain-of-X methods for LLMs
• Categorize CoX by nodes and tasks
• Discuss findings and future directions

Innovation

• Explores Chain-of-X methods
• Categorizes by nodes and tasks
• Surveys applications for LLMs