Understanding The Robustness of Self-supervised Learning Through Topic Modeling
This addresses the robustness issue in topic modeling for NLP researchers, providing insights into self-supervised learning's advantages, though it is incremental as it builds on existing methods in a specific context.
The paper tackles the problem of understanding why self-supervised learning outperforms traditional methods like probabilistic models in topic modeling, showing that it can recover useful posterior information without being sensitive to model misspecification, with empirical results indicating performance on par with correct model inference and better than misspecified models.
Self-supervised learning has significantly improved the performance of many NLP tasks. However, how can self-supervised learning discover useful representations, and why is it better than traditional approaches such as probabilistic models are still largely unknown. In this paper, we focus on the context of topic modeling and highlight a key advantage of self-supervised learning - when applied to data generated by topic models, self-supervised learning can be oblivious to the specific model, and hence is less susceptible to model misspecification. In particular, we prove that commonly used self-supervised objectives based on reconstruction or contrastive samples can both recover useful posterior information for general topic models. Empirically, we show that the same objectives can perform on par with posterior inference using the correct model, while outperforming posterior inference using misspecified models.