Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fine-tuning vs Prompting, Can Language Models Understand Human Values?

Pingwei Sun

TL;DR

The potential of fine-tuning and prompt tuning in this downstream task, using the Human Value Detection 2023, is explored and whether models can effectively solve the problem based on the knowledge acquired during the pre-training stage is validated.

Abstract

Accurately handling the underlying support values in sentences is crucial for understanding the speaker's tendencies, yet it poses a challenging task in natural language understanding (NLU). In this article, we explore the potential of fine-tuning and prompt tuning in this downstream task, using the Human Value Detection 2023. Additionally, we attempt to validate whether models can effectively solve the problem based on the knowledge acquired during the pre-training stage. Simultaneously, our interest lies in the capabilities of large language models (LLMs) aligned with RLHF in this task, and some preliminary attempts are presented.

Fine-tuning vs Prompting, Can Language Models Understand Human Values?

TL;DR

The potential of fine-tuning and prompt tuning in this downstream task, using the Human Value Detection 2023, is explored and whether models can effectively solve the problem based on the knowledge acquired during the pre-training stage is validated.

Abstract

Accurately handling the underlying support values in sentences is crucial for understanding the speaker's tendencies, yet it poses a challenging task in natural language understanding (NLU). In this article, we explore the potential of fine-tuning and prompt tuning in this downstream task, using the Human Value Detection 2023. Additionally, we attempt to validate whether models can effectively solve the problem based on the knowledge acquired during the pre-training stage. Simultaneously, our interest lies in the capabilities of large language models (LLMs) aligned with RLHF in this task, and some preliminary attempts are presented.
Paper Structure (19 sections, 2 equations, 2 figures, 5 tables)

This paper contains 19 sections, 2 equations, 2 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methods
  4. Data Utilization
  5. Fine-tuning as a Classification Task
  6. Contrastive Learning
  7. Prompt Tuning Techniques
  8. Experiments
  9. Exp-I: Fine-tuning
  10. Exp-II: Prompt Tuning
  11. Exp-III: Works on LLMs
  12. Analysis
  13. Effectiveness of Fine-tuning and Input Description
  14. Performance of Prompt Tuning
  15. Prompting LLMs to Detect Human Values
  16. ...and 4 more sections

Figures (2)

  • Figure 1: Illustration of the workflow of our experiments. Tracks marked by different colors stand for combinations of models, tasks, and methods, details of which are told in § \ref{['Sec4']}. The dashed lines mean methods are theoretically available but we do not consider them in the project because of poor performance or being computationally demanding. The colorful dots are experiment results and indicate the analysis for the questions proposed in § \ref{['Sec1']}.
  • Figure 2: Prompt templates for different task processing modes, including classification(CLS), masked binary choice(MBC), binary choice answering(BCA), open answering(OA), and Chain-of-Thought(CoT). The bolded content in brackets represents the features in the dataset, shown in Table \ref{['Tab1']}.