Can Unconditional Language Models Recover Arbitrary Sentences?

By Nishant Subramani, Samuel R. Bowman, Kyunghyun Cho. 2019

29 citations recorded by the source corpus.

Neural network-based generative language models like ELMo and BERT can work effectively as general purpose sentence encoders in text classification without further fine-tuning. Is it possible to adapt them in a similar way for use as general-purpose decoders? For this to be possible, it would need to be the case that for any target sentence of interest, there is some continuous representation that can be passed to the language model to cause it to reproduce that sentence. We set aside the difficult problem of designing an encoder that can produce such representations and instead ask directly whether such representations exist at all. To do this, we introduce a pair of effective complementary methods for feeding representations into pretrained unconditional language models and a corresponding set of methods to map sentences into and out of this representation space, the \textit{reparametrized sentence space}. We then investigate the conditions under which a language model can be made to generate a sentence through the identification of a point in such a space and find that it is possible to recover arbitrary sentences nearly perfectly with language models and representations of moderate size.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Large Language Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Large Language Models

Citations in the corpus

Prefix-Tuning: Optimizing Continuous Prompts for Generation (2021, methodology): “Continuous vectors have been used to steer LMs; for example, Subramani et al. (2020) showed that a pretrained LSTM language model can reconstruct arbitrary sentences by optimizing a continuous vector for each sentence, making the vector inputspecific.”
Plug and Play Language Models: A Simple Approach to Controlled Text Generation (2019, background): “In another interesting but tangential piece of work, Subramani et al. (2019) recently showed that a pre-trained language model can be steered to recover arbitrary sentences.”
Sparse, Dense, and Attentional Representations for Text Retrieval (2020, background): “Nonetheless, (Subramani et al., 2019) empirically observe a similar dependence between sentence length and embedding size.”
FUDGE: Controlled Text Generation With Future Discriminators (2021, background): “In fact, Subramani et al. (2019) show that by carefully modifying the latent state, one can cause the base G to produce arbitrary outputs.”
Plug-and-Play Conversational Models (2020, methodology): “Controlled Text Generation Recent methods for controlled generation include fine-tuning models using supervised learning (Peng et al., 2020; Subramani et al., 2019), reinforcement learning (Ziegler et al.”
Don’t Say What You Don’t Know: Improving the Consistency of Abstractive Summarization by Constraining Beam Search (2022, background): “Pretrained language models are capable of producing incredibly fluent text and prior work on steering them over-optimizes for maximizing the highest likelihood output (Subramani et al., 2019; Subramani and Suresh, 2020).”
Conditioned Natural Language Generation using only Unconditioned Language Model: An Exploration (2020, background): “In (Subramani et al., 2019), the authors show that a pre-trained language model can be steered to recover arbitrary sentences.”