ExpeTrans: LLMs Are Experiential Transfer Learners

2025-05-29

AI Summary

How can large language models (LLMs) autonomously transfer cross-task experience to reduce reliance on manual annotation and expert-curated demonstrations? This paper proposes ExpeTrans, the first end-to-end framework enabling LLMs to automatically extract, model task similarity, remap, and adaptively inject textual reasoning experiences from source tasks into novel target tasks—without human-designed prompts. ExpeTrans breaks from conventional prompt-engineering paradigms by emulating human-like generalization through cognitive-inspired experience abstraction and transfer. Evaluated on 13 diverse benchmark datasets, it consistently improves LLM performance across tasks. Ablation studies validate the necessity and efficacy of each module, while interpretability analyses confirm the plausibility and mechanism of experience transfer. The core contributions are threefold: (1) the first fully autonomous, end-to-end experience transfer framework for LLMs; (2) seamless integration of task-agnostic generalization, full automation, and transparency; and (3) empirical demonstration of robust, interpretable cross-task knowledge reuse.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Computer Vision: Large Vision Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Recent studies provide large language models (LLMs) with textual task-solving experiences via prompts to improve their performance. However, previous methods rely on substantial human labor or time to gather such experiences for each task, which is impractical given the growing variety of task types in user queries to LLMs. To address this issue, we design an autonomous experience transfer framework to explore whether LLMs can mimic human cognitive intelligence to autonomously transfer experience from existing source tasks to newly encountered target tasks. This not only allows the acquisition of experience without extensive costs of previous methods, but also offers a novel path for the generalization of LLMs. Experimental results on 13 datasets demonstrate that our framework effectively improves the performance of LLMs. Furthermore, we provide a detailed analysis of each module in the framework.

Problem

• Autonomous transfer of task-solving experiences to new tasks
• Reducing human effort in gathering LLM training experiences
• Improving LLM generalization across diverse task types

Innovation

• Autonomous transfer of task-solving experiences
• Mimics human cognitive intelligence
• Improves LLM performance without human labor