Pre-training Text-to-Text Transformers for Concept-centric Common Sense
Pre-trained language models (PTLM) have achieved impressive results in a range of natural language understanding (NLU) and generation (NLG) tasks. However, current pre-training objectives such as masked token prediction (for BERT-style PTLMs) and masked span infilling (for T5-style PTLMs) do not explicitly model the relational commonsense knowledge about everyday concepts, which is crucial to many downstream tasks that need common sense to understand or generate. To augment PTLMs with concept-centric commonsense knowledge, in this paper, we propose both generative and contrastive objectives for learning common sense from the text, and use them as intermediate self-supervised learning tasks for incrementally pre-training PTLMs (before task-specific fine-tuning on downstream datasets). Furthermore, we develop a joint pre-training framework to unify generative and contrastive objectives so that they can mutually reinforce each other. Extensive experimental results show that our method, concept-aware language model (CALM), can pack more commonsense knowledge into the parameters of a pre-trained text-to-text transformer without relying on external knowledge graphs, yielding better performance on both NLU and NLG tasks. We show that while only incrementally pre-trained on a relatively small corpus for a few steps, CALM outperforms baseline methods by a consistent margin and even comparable with some larger PTLMs, which suggests that CALM can serve as a general, plug-and-play method for improving the commonsense reasoning ability of a PTLM.
Code (1)
Tasks
Common Sense ReasoningKnowledge GraphsLanguage ModellingNatural Language UnderstandingSelf-Supervised LearningSimilar Papers 제목 키워드 기반
Concept-Centric Transformers: Enhancing Model Interpretability through Object-Centric Concept Learning within a Shared Global Workspace
Many interpretable AI approaches have been proposed to provide plausible explanations for a model's decision-making. However, configuring an explainable model that effectively communicates among computational modules has…
ClassificationDecision MakingLayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking
Self-supervised pre-training techniques have achieved remarkable progress in Document AI. Most multimodal pre-trained models use a masked language modeling objective to learn bidirectional representations on the text mod…
cross-modal alignmentDocument AIdocument-image-classificationDocument Image Classification+16Transformers are Multi-State RNNs
Transformers are considered conceptually different from the previous generation of state-of-the-art NLP models - recurrent neural networks (RNNs). In this work, we demonstrate that decoder-only transformers can in fact b…
DecoderEgoNCE++: Do Egocentric Video-Language Models Really Understand Hand-Object Interactions?
Egocentric video-language pretraining is a crucial paradigm to advance the learning of egocentric hand-object interactions (EgoHOI). Despite the great success on existing testbeds, these benchmarks focus more on closed-s…
Action RecognitionAttributeIn-Context LearningMulti-Instance RetrievalThe Coverage Principle: A Framework for Understanding Compositional Generalization
Large language models excel at pattern matching, yet often fall short in systematic compositional generalization. We propose the coverage principle: a data-centric framework showing that models relying primarily on patte…