SideControl: Controlled Open-domain Dialogue Generation via Additive Side Networks

Abstract

Transformer-based pre-trained language models boost the performance of open-domain dialogue systems. Prior works leverage Transformer-based pre-trained language models to generate texts with desired attributes in two general approaches: (1) gradient-based methods: updating all latent representations of pre-trained models with gradients from attribute models; (2) weighted-decoding methods: re-ranking beam candidates from pre-trained models with attribute functions. However, gradient-based methods lead to high computation cost and can easily get overfitted on small training sets, while weighted-decoding methods are inherently constrained by the low-variance high-bias pre-trained model. In this work, we propose a novel approach to control the generation of Transformer-based pre-trained language models: the SideControl framework, which leverages a novel control attributes loss to incorporate useful control signals, and is shown to perform well with very limited training samples. We evaluate our proposed method on two benchmark open-domain dialogue datasets, and results show that the SideControl framework has better controllability, higher generation quality and better sample-efficiency than existing gradient-based and weighted-decoding baselines.

arXiv commentEMNLP2021 (Findings)

References in corpus (8)

• Adam: A Method for Stochastic Optimization
• Neural Machine Translation by Jointly Learning to Align and Translate
• Decoupled Weight Decay Regularization
• CTRL: A Conditional Transformer Language Model for Controllable Generation
• DailyDialog: A Manually Labelled Multi-turn Dialogue Dataset
• Wizard of Wikipedia: Knowledge-Powered Conversational agents
• Fine-Tuning Language Models from Human Preferences
• ACUTE-EVAL: Improved Dialogue Evaluation with Optimized Questions and Multi-turn Comparisons

Cited by in corpus (1)

• Few-Shot Bot: Prompt-Based Learning for Dialogue Systems