Mockingbird: How does LLM perform in general machine learning tasks?

TL;DR

This paper introduces Mockingbird, a framework that adapts large language models to general machine learning tasks by instructing them to role-play and self-reflect, demonstrating promising but limited results compared to domain-specific feedback.

Contribution

The paper presents a novel framework, Mockingbird, for leveraging LLMs in general machine learning tasks through role-playing and self-reflection techniques.

Findings

LLMs can perform general machine learning tasks with acceptable results.

Self-reflection alone does not surpass domain-specific feedback.

Mockingbird demonstrates scalability across multiple tasks.

Abstract

Large language models (LLMs) are now being used with increasing frequency as chat bots, tasked with the summarizing information or generating text and code in accordance with user instructions. The rapid increase in reasoning capabilities and inference speed of LLMs has revealed their remarkable potential for applications extending beyond the domain of chat bots to general machine learning tasks. This work is conducted out of the curiosity about such potential. In this work, we propose a framework Mockingbird to adapt LLMs to general machine learning tasks and evaluate its performance and scalability on several general machine learning tasks. The core concept of this framework is instructing LLMs to role-play functions and reflect on its mistakes to improve itself. Our evaluation and analysis result shows that LLM-driven machine learning methods, such as Mockingbird, can achieve acceptable results on common machine learning tasks; however, solely reflecting on its own currently cannot outperform the effect of domain-specific documents and feedback from human experts.