Critic-Guided Decoding for Controlled Text Generation

DOI

10.18653/v1/2023.findings-acl.281

Abstract

Steering language generation towards objectives or away from undesired content has been a long-standing goal in utilizing language models (LM). Recent work has demonstrated reinforcement learning and weighted decoding as effective approaches to achieve a higher level of language control and quality with pros and cons. In this work, we propose a novel critic decoding method for controlled language generation (CriticControl) that combines the strengths of reinforcement learning and weighted decoding. Specifically, we adopt the actor-critic framework and train an LM-steering critic from reward models. Similar to weighted decoding, our method freezes the language model and manipulates the output token distribution using a critic to improve training efficiency and stability. Evaluation of our method on three controlled generation tasks, topic control, sentiment control, and detoxification, shows that our approach generates more coherent and well-controlled texts than previous methods. In addition, CriticControl demonstrates superior generalization ability in zero-shot settings. Human evaluation studies also corroborate our findings.

Document Type

Conference Proceeding

Publication Date

7-2023

Publisher Statement

ACL materials are Copyright © 1963–2024 ACL; other materials are copyrighted by their respective copyright holders. Materials prior to 2016 here are licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 International License. Permission is granted to make copies for the purposes of teaching and research. Materials published in or after 2016 are licensed on a Creative Commons Attribution 4.0 International License.

Link to Full Text

Share

COinS