Large-scale pre-trained language models have demonstrated strong capabilities of generating realistic text. However, it remains challenging to control the generation results. Previous approaches such as prompting are far from sufficient, which limits the usage of language models. To tackle this challenge, we propose an innovative method, inverse prompting, to better control text generation. The core idea of inverse prompting is to use generated text to inversely predict the prompt during beam search, which enhances the relevance between the prompt and the generated text and provides better controllability. Empirically, we pre-train a large-scale Chinese language model to perform a systematic study using human evaluation on the tasks of open-domain poem generation and open-domain long-form question answering. Our results show that our proposed method substantially outperforms the baselines and that our generation quality is close to human performance on some of the tasks. Narrators can try our poem generation demo at https://pretrain.aminer.cn/apps/poetry.html, while our QA demo can be found at https://pretrain.aminer.cn/app/qa. For researchers, the code is provided in https://github.com/THUDM/InversePrompting.
To enrich language models with domain knowledge is crucial but difficult. Based on the world's largest public academic graph Open Academic Graph (OAG), we pre-train an academic language model, namely OAG-BERT, which integrates massive heterogeneous entities including paper, author, concept, venue, and affiliation. To better endow OAG-BERT with the ability to capture entity information, we develop novel pre-training strategies including heterogeneous entity type embedding, entity-aware 2D positional encoding, and span-aware entity masking. For zero-shot inference, we design a special decoding strategy to allow OAG-BERT to generate entity names from scratch. We evaluate the OAG-BERT on various downstream academic tasks, including NLP benchmarks, zero-shot entity inference, heterogeneous graph link prediction, and author name disambiguation. Results demonstrate the effectiveness of the proposed pre-training approach to both comprehending academic texts and modeling knowledge from heterogeneous entities. OAG-BERT has been deployed to multiple real-world applications, such as reviewer recommendations for NSFC (National Nature Science Foundation of China) and paper tagging in the AMiner system. It is also available to the public through the CogDL package.
We propose SentiBERT, a variant of BERT that effectively captures compositional sentiment semantics. The model incorporates contextualized representation with binary constituency parse tree to capture semantic composition. Comprehensive experiments demonstrate that SentiBERT achieves competitive performance on phrase-level sentiment classification. We further demonstrate that the sentiment composition learned from the phrase-level annotations on SST can be transferred to other sentiment analysis tasks as well as related tasks, such as emotion classification tasks. Moreover, we conduct ablation studies and design visualization methods to understand SentiBERT. We show that SentiBERT is better than baseline approaches in capturing negation and the contrastive relation and model the compositional sentiment semantics.
We propose VisualBERT, a simple and flexible framework for modeling a broad range of vision-and-language tasks. VisualBERT consists of a stack of Transformer layers that implicitly align elements of an input text and regions in an associated input image with self-attention. We further propose two visually-grounded language model objectives for pre-training VisualBERT on image caption data. Experiments on four vision-and-language tasks including VQA, VCR, NLVR2, and Flickr30K show that VisualBERT outperforms or rivals with state-of-the-art models while being significantly simpler. Further analysis demonstrates that VisualBERT can ground elements of language to image regions without any explicit supervision and is even sensitive to syntactic relationships, tracking, for example, associations between verbs and image regions corresponding to their arguments.