Several multi-hop reading comprehension datasets have been proposed to resolve the issue of reasoning shortcuts by which questions can be answered without performing multi-hop reasoning. However, the ability of multi-hop models to perform step-by-step reasoning when finding an answer to a comparison question remains unclear. It is also unclear how questions about the internal reasoning process are useful for training and evaluating question-answering (QA) systems. To evaluate the model precisely in a hierarchical manner, we first propose a dataset, \textit{HieraDate}, with three probing tasks in addition to the main question: extraction, reasoning, and robustness. Our dataset is created by enhancing two previous multi-hop datasets, HotpotQA and 2WikiMultiHopQA, focusing on multi-hop questions on date information that involve both comparison and numerical reasoning. We then evaluate the ability of existing models to understand date information. Our experimental results reveal that the multi-hop models do not have the ability to subtract two dates even when they perform well in date comparison and number subtraction tasks. Other results reveal that our probing questions can help to improve the performance of the models (e.g., by +10.3 F1) on the main QA task and our dataset can be used for data augmentation to improve the robustness of the models.

Question Answering is a popular application of NLP. Transformer models trained on big datasets have dramatically improved the state-of-the-art results on Question Answering. The question answering task can be formulated in many ways. The most common application is an extractive question answering on a small context. The SQuAD dataset is a popular dataset where given a passage and a question, the model selects the word(s) representing the answer.

With the increasing importance of safety requirements associated with the use of black box models, evaluation of selective answering capability of models has been critical. Area under the curve (AUC) is used as a metric for this purpose. We find limitations in AUC; e.g., a model having higher AUC is not always better in performing selective answering. We propose three alternate metrics that fix the identified limitations. On experimenting with ten models, our results using the new metrics show that newer and larger pre-trained models do not necessarily show better performance in selective answering. We hope our insights will help develop better models tailored for safety-critical applications.

Models for Visual Question Answering (VQA) often rely on the spurious correlations, i.e., the language priors, that appear in the biased samples of training set, which make them brittle against the out-of-distribution (OOD) test data. Recent methods have achieved promising progress in overcoming this problem by reducing the impact of biased samples on model training. However, these models reveal a trade-off that the improvements on OOD data severely sacrifice the performance on the in-distribution (ID) data (which is dominated by the biased samples). Therefore, we propose a novel contrastive learning approach, MMBS, for building robust VQA models by Making the Most of Biased Samples. Specifically, we construct positive samples for contrastive learning by eliminating the information related to spurious correlation from the original training samples and explore several strategies to use the constructed positive samples for training. Instead of undermining the importance of biased samples in model training, our approach precisely exploits the biased samples for unbiased information that contributes to reasoning. The proposed method is compatible with various VQA backbones. We validate our contributions by achieving competitive performance on the OOD dataset VQA-CP v2 while preserving robust performance on the ID dataset VQA v2.

Conversational question generation (CQG) serves as a vital task for machines to assist humans, such as interactive reading comprehension, through conversations. Compared to traditional single-turn question generation (SQG), CQG is more challenging in the sense that the generated question is required not only to be meaningful, but also to align with the occurred conversation history. While previous studies mainly focus on how to model the flow and alignment of the conversation, there has been no thorough study to date on which parts of the context and history are necessary for the model. We argue that shortening the context and history is crucial as it can help the model to optimise more on the conversational alignment property. To this end, we propose CoHS-CQG, a two-stage CQG framework, which adopts a CoHS module to shorten the context and history of the input. In particular, CoHS selects contiguous sentences and history turns according to their relevance scores by a top-p strategy. Our model achieves state-of-the-art performances on CoQA in both the answer-aware and answer-unaware settings.

Question Generation (QG) aims to automate the task of composing questions for a passage with a set of chosen answers found within the passage. In recent years, the introduction of neural generation models has resulted in substantial improvements of automatically generated questions in terms of quality, especially compared to traditional approaches that employ manually crafted heuristics. However, the metrics commonly applied in QG evaluations have been criticized for their low agreement with human judgement. We therefore propose a new reference-free evaluation metric that has the potential to provide a better mechanism for evaluating QG systems, called QAScore. Instead of fine-tuning a language model to maximize its correlation with human judgements, QAScore evaluates a question by computing the cross entropy according to the probability that the language model can correctly generate the masked words in the answer to that question. Furthermore, we conduct a new crowd-sourcing human evaluation experiment for the QG evaluation to investigate how QAScore and other metrics can correlate with human judgements. Experiments show that QAScore obtains a stronger correlation with the results of our proposed human evaluation method compared to existing traditional word-overlap-based metrics such as BLEU and ROUGE, as well as the existing pretrained-model-based metric BERTScore.

Generative question answering (QA) models generate answers to questions either solely based on the parameters of the model (the closed-book setting) or additionally retrieving relevant evidence (the open-book setting). Generative QA models can answer some relatively complex questions, but the mechanism through which they do so is still poorly understood. We perform several studies aimed at better understanding the multi-hop reasoning capabilities of generative QA models. First, we decompose multi-hop questions into multiple corresponding single-hop questions, and find marked inconsistency in QA models' answers on these pairs of ostensibly identical question chains. Second, we find that models lack zero-shot multi-hop reasoning ability: when trained only on single-hop questions, models generalize poorly to multi-hop questions. Finally, we demonstrate that it is possible to improve models' zero-shot multi-hop reasoning capacity through two methods that approximate real multi-hop natural language (NL) questions by training on either concatenation of single-hop questions or logical forms (SPARQL). In sum, these results demonstrate that multi-hop reasoning does not emerge naturally in generative QA models, but can be encouraged by advances in training or modeling techniques.

This paper addresses the task of conversational question answering (ConvQA) over knowledge graphs (KGs). The majority of existing ConvQA methods rely on full supervision signals with a strict assumption of the availability of gold logical forms of queries to extract answers from the KG. However, creating such a gold logical form is not viable for each potential question in a real-world scenario. Hence, in the case of missing gold logical forms, the existing information retrieval-based approaches use weak supervision via heuristics or reinforcement learning, formulating ConvQA as a KG path ranking problem. Despite missing gold logical forms, an abundance of conversational contexts, such as entire dialog history with fluent responses and domain information, can be incorporated to effectively reach the correct KG path. This work proposes a contrastive representation learning-based approach to rank KG paths effectively. Our approach solves two key challenges. Firstly, it allows weak supervision-based learning that omits the necessity of gold annotations. Second, it incorporates the conversational context (entire dialog history and domain information) to jointly learn its homogeneous representation with KG paths to improve contrastive representations for effective path ranking. We evaluate our approach on standard datasets for ConvQA, on which it significantly outperforms existing baselines on all domains and overall. Specifically, in some cases, the Mean Reciprocal Rank (MRR) and Hit@5 ranking metrics improve by absolute 10 and 18 points, respectively, compared to the state-of-the-art performance.

With the recent advance in large pre-trained language models, researchers have achieved record performances in NLP tasks that mostly focus on language pattern matching. The community is experiencing the shift of the challenge from how to model language to the imitation of complex reasoning abilities like human beings. In this work, we investigate the application domain of finance that involves real-world, complex numerical reasoning. We propose a new large-scale dataset, ConvFinQA, aiming to study the chain of numerical reasoning in conversational question answering. Our dataset poses great challenge in modeling long-range, complex numerical reasoning paths in real-world conversations. We conduct comprehensive experiments and analyses with both the neural symbolic methods and the prompting-based methods, to provide insights into the reasoning mechanisms of these two divisions. We believe our new dataset should serve as a valuable resource to push forward the exploration of real-world, complex reasoning tasks as the next research focus. Our dataset and code is publicly available at https://github.com/czyssrs/ConvFinQA.

Embodied agents are expected to perform more complicated tasks in an interactive environment, with the progress of Embodied AI in recent years. Existing embodied tasks including Embodied Referring Expression (ERE) and other QA-form tasks mainly focuses on interaction in term of linguistic instruction. Therefore, enabling the agent to manipulate objects in the environment for exploration actively has become a challenging problem for the community. To solve this problem, We introduce a new embodied task: Remote Embodied Manipulation Question Answering (REMQA) to combine ERE with manipulation tasks. In the REMQA task, the agent needs to navigate to a remote position and perform manipulation with the target object to answer the question. We build a benchmark dataset for the REMQA task in the AI2-THOR simulator. To this end, a framework with 3D semantic reconstruction and modular network paradigms is proposed. The evaluation of the proposed framework on the REMQA dataset is presented to validate its effectiveness.