Controllable Natural Language Generation with Contrastive Prefixes
This work addresses the need for efficient and effective control over text generation in NLP applications, representing an incremental improvement over existing prefix-based methods.
The authors tackled the problem of controllable natural language generation with large pretrained language models by proposing a lightweight framework using attribute-specific vectors called prefixes, achieving results that guide generation towards desired attributes while maintaining high linguistic quality.
To guide the generation of large pretrained language models (LM), previous work has focused on directly fine-tuning the language model or utilizing an attribute discriminator. In this work, we propose a novel lightweight framework for controllable GPT2 generation, which utilizes a set of small attribute-specific vectors, called prefixes, to steer natural language generation. Different from prefix-tuning, where each prefix is trained independently, we take the relationship among prefixes into consideration and train multiple prefixes simultaneously. We propose a novel supervised method and also an unsupervised method to train the prefixes for single-aspect control while the combination of these two methods can achieve multi-aspect control. Experimental results on both single-aspect and multi-aspect control show that our methods can guide generation towards the desired attributes while keeping high linguistic quality.