Transformer-based language models have been shown to be highly effective for several NLP tasks. In this article, we consider three transformer models, BERT, RoBERTa, and XLNet, in both small and large versions, and investigate how faithful their representations are with respect to the semantic content of texts. We formalize a notion of semantic faithfulness, in which the semantic content of a text should causally figure in a model’s inferences in question answering. We then test this notion by observing a model’s behavior on answering questions about a story after performing two novel semantic interventions—deletion intervention and negation intervention. While transformer models achieve high performance on standard question answering tasks, we show that they fail to be semantically faithful once we perform these interventions for a significant number of cases (∼ 50% for deletion intervention, and ∼ 20% drop in accuracy for negation intervention). We then propose an intervention-based training regime that can mitigate the undesirable effects for deletion intervention by a significant margin (from ∼ 50% to ∼ 6%). We analyze the inner-workings of the models to better understand the effectiveness of intervention-based training for deletion intervention. But we show that this training does not attenuate other aspects of semantic unfaithfulness such as the models’ inability to deal with negation intervention or to capture the predicate–argument structure of texts. We also test InstructGPT, via prompting, for its ability to handle the two interventions and to capture predicate–argument structure. While InstructGPT models do achieve very high performance on predicate–argument structure task, they fail to respond adequately to our deletion and negation interventions.

With the advent of large language models (LLMs), the trend in NLP has been to train LLMs on vast amounts of data to solve diverse language understanding and generation tasks. The list of LLM successes is long and varied. Nevertheless, several recent papers provide empirical evidence that LLMs fail to capture important aspects of linguistic meaning. Focusing on universal quantification, we provide a theoretical foundation for these empirical findings by proving that LLMs cannot learn certain fundamental semantic properties including semantic entailment and consistency as they are defined in formal semantics. More generally, we show that LLMs are unable to learn concepts beyond the first level of the Borel Hierarchy, which imposes severe limits on the ability of LMs, both large and small, to capture many aspects of linguistic meaning. This means that LLMs will operate without formal guarantees on tasks that require entailments and deep linguistic understanding.

The task of Question Answering is at the very core of machine comprehension. In this paper, we propose a Convolutional Neural Network (CNN) model for text-based multiple choice question answering where questions are based on a particular article. Given an article and a multiple choice question, our model assigns a score to each question-option tuple and chooses the final option accordingly. We test our model on Textbook Question Answering (TQA) and SciQ dataset. Our model outperforms several LSTM-based baseline models on the two datasets.

We propose a novel neural lemmatization model which is language independent and supervised in nature. To handle the words in a neural framework, word embedding technique is used to represent words as vectors. The proposed lemmatizer makes use of contextual information of the surface word to be lemmatized. Given a word along with its contextual neighbours as input, the model is designed to produce the lemma of the concerned word as output. We introduce a new network architecture that permits only dimension specific connections between the input and the output layer of the model. For the present work, Bengali is taken as the reference language. Two datasets are prepared for training and testing purpose consisting of 19,159 and 2,126 instances respectively. As Bengali is a resource scarce language, these datasets would be beneficial for the respective research community. Evaluation method shows that the neural lemmatizer achieves 69.57% accuracy on the test dataset and outperforms the simple cosine similarity based baseline strategy by a margin of 1.37%.