While deep learning models are making fast progress on the task of Natural Language Inference, recent studies have also shown that these models achieve high accuracy by exploiting several dataset biases, and without deep understanding of the language semantics. Using contradiction-word bias and word-overlapping bias as our two bias examples, this paper explores both data-level and model-level debiasing methods to robustify models against lexical dataset biases. First, we debias the dataset through data augmentation and enhancement, but show that the model bias cannot be fully removed via this method. Next, we also compare two ways of directly debiasing the model without knowing what the dataset biases are in advance. The first approach aims to remove the label bias at the embedding level. The second approach employs a bag-of-words sub-model to capture the features that are likely to exploit the bias and prevents the original model from learning these biased features by forcing orthogonality between these two sub-models. We performed evaluations on new balanced datasets extracted from the original MNLI dataset as well as the NLI stress tests, and show that the orthogonality approach is better at debiasing the model while maintaining competitive overall accuracy. Our code and data are available at: https://github.com/owenzx/LexicalDebias-ACL2020
Videos convey rich information. Dynamic spatio-temporal relationships between people/objects, and diverse multimodal events are present in a video clip. Hence, it is important to develop automated models that can accurately extract such information from videos. Answering questions on videos is one of the tasks which can evaluate such AI abilities. In this paper, we propose a video question answering model which effectively integrates multi-modal input sources and finds the temporally relevant information to answer questions. Specifically, we first employ dense image captions to help identify objects and their detailed salient regions and actions, and hence give the model useful extra information (in explicit textual format to allow easier matching) for answering questions. Moreover, our model is also comprised of dual-level attention (word/object and frame level), multi-head self/cross-integration for different sources (video and dense captions), and gates which pass more relevant information to the classifier. Finally, we also cast the frame selection problem as a multi-label classification task and introduce two loss functions, In-andOut Frame Score Margin (IOFSM) and Balanced Binary Cross-Entropy (BBCE), to better supervise the model with human importance annotations. We evaluate our model on the challenging TVQA dataset, where each of our model components provides significant gains, and our overall model outperforms the state-of-the-art by a large margin (74.09% versus 70.52%). We also present several word, object, and frame level visualization studies. Our code is publicly available at: https://github.com/hyounghk/VideoQADenseCapFrameGate-ACL2020
Generating multi-sentence descriptions for videos is one of the most challenging captioning tasks due to its high requirements for not only visual relevance but also discourse-based coherence across the sentences in the paragraph. Towards this goal, we propose a new approach called Memory-Augmented Recurrent Transformer (MART), which uses a memory module to augment the transformer architecture. The memory module generates a highly summarized memory state from the video segments and the sentence history so as to help better prediction of the next sentence (w.r.t. coreference and repetition aspects), thus encouraging coherent paragraph generation. Extensive experiments, human evaluations, and qualitative analyses on two popular datasets ActivityNet Captions and YouCookII show that MART generates more coherent and less repetitive paragraph captions than baseline methods, while maintaining relevance to the input video events. All code is available open-source at: https://github.com/jayleicn/recurrent-transformer
Vision-and-Language Navigation (VLN) requires an agent to follow natural-language instructions, explore the given environments, and reach the desired target locations. These step-by-step navigational instructions are crucial when the agent is navigating new environments about which it has no prior knowledge. Most recent works that study VLN observe a significant performance drop when tested on unseen environments (i.e., environments not used in training), indicating that the neural agent models are highly biased towards training environments. Although this issue is considered as one of the major challenges in VLN research, it is still under-studied and needs a clearer explanation. In this work, we design novel diagnosis experiments via environment re-splitting and feature replacement, looking into possible reasons for this environment bias. We observe that neither the language nor the underlying navigational graph, but the low-level visual appearance conveyed by ResNet features directly affects the agent model and contributes to this environment bias in results. According to this observation, we explore several kinds of semantic representations that contain less low-level visual information, hence the agent learned with these features could be better generalized to unseen testing environments. Without modifying the baseline agent model and its training method, our explored semantic features significantly decrease the performance gaps between seen and unseen on multiple datasets (i.e. R2R, R4R, and CVDN) and achieve competitive unseen results to previous state-of-the-art models. Our code and features are available at: https://github.com/zhangybzbo/EnvBiasVLN
Algorithmic approaches to interpreting machine learning models have proliferated in recent years. We carry out human subject tests that are the first of their kind to isolate the effect of algorithmic explanations on a key aspect of model interpretability, simulatability, while avoiding important confounding experimental factors. A model is simulatable when a person can predict its behavior on new inputs. Through two kinds of simulation tests involving text and tabular data, we evaluate five explanations methods: (1) LIME, (2) Anchor, (3) Decision Boundary, (4) a Prototype model, and (5) a Composite approach that combines explanations from each method. Clear evidence of method effectiveness is found in very few cases: LIME improves simulatability in tabular classification, and our Prototype method is effective in counterfactual simulation tests. We also collect subjective ratings of explanations, but we do not find that ratings are predictive of how helpful explanations are. Our results provide the first reliable and comprehensive estimates of how explanations influence simulatability across a variety of explanation methods and data domains. We show that (1) we need to be careful about the metrics we use to evaluate explanation methods, and (2) there is significant room for improvement in current methods. All our supporting code, data, and models are publicly available at: https://github.com/peterbhase/InterpretableNLP-ACL2020
Reading comprehension models often overfit to nuances of training datasets and fail at adversarial evaluation. Training with adversarially augmented dataset improves robustness against those adversarial attacks but hurts generalization of the models. In this work, we present several effective adversaries and automated data augmentation policy search methods with the goal of making reading comprehension models more robust to adversarial evaluation, but also improving generalization to the source domain as well as new domains and languages. We first propose three new methods for generating QA adversaries, that introduce multiple points of confusion within the context, show dependence on insertion location of the distractor, and reveal the compounding effect of mixing adversarial strategies with syntactic and semantic paraphrasing methods. Next, we find that augmenting the training datasets with uniformly sampled adversaries improves robustness to the adversarial attacks but leads to decline in performance on the original unaugmented dataset. We address this issue via RL and more efficient Bayesian policy search methods for automatically learning the best augmentation policy combinations of the transformation probability for each adversary in a large search space. Using these learned policies, we show that adversarial training can lead to significant improvements in in-domain, out-of-domain, and cross-lingual (German, Russian, Turkish) generalization without any use of training data from the target domain or language.
We find that the performance of state-of-the-art models on Natural Language Inference (NLI) and Reading Comprehension (RC) analysis/stress sets can be highly unstable. This raises three questions: (1) How will the instability affect the reliability of the conclusions drawn based on these analysis sets? (2) Where does this instability come from? (3) How should we handle this instability and what are some potential solutions? For the first question, we conduct a thorough empirical study over analysis sets and find that in addition to the unstable final performance, the instability exists all along the training curve. We also observe lower-than-expected correlations between the analysis validation set and standard validation set, questioning the effectiveness of the current model-selection routine. Next, to answer the second question, we give both theoretical explanations and empirical evidence regarding the source of the instability, demonstrating that the instability mainly comes from high inter-example correlations within analysis sets. Finally, for the third question, we discuss an initial attempt to mitigate the instability and suggest guidelines for future work such as reporting the decomposed variance for more interpretable results and fair comparison across models. Our code is publicly available at: https://github. com/owenzx/InstabilityAnalysis
We introduce a new multimodal retrieval task - TV show Retrieval (TVR), in which a short video moment has to be localized from a large video (with subtitle) corpus, given a natural language query. Different from previous moment retrieval tasks dealing with videos only, TVR requires the system to understand both the video and the associated subtitle text, making it a more realistic task. To support the study of this new task, we have collected a large-scale, high-quality dataset consisting of 108,965 queries on 21,793 videos from 6 TV shows of diverse genres, where each query is associated with a tight temporal alignment. Strict qualification and post-annotation verification tests are applied to ensure the quality of the collected data. We present several baselines and a novel Cross-modal Moment Localization (XML) modular network for this new dataset and task. The proposed XML model surpasses all presented baselines by a large margin and with better efficiency, providing a strong starting point for future work. Extensive analysis experiments also show that incorporating both video and subtitle modules yields better performance than either alone. Lastly, we have also collected additional descriptions for each annotated moment in TVR to form a new multimodal captioning dataset with 262K captions, named the TV show Caption dataset (TVC). Here models need to jointly use the video and subtitle to generate a caption description. Both datasets are publicly available at https://tvr.cs.unc.edu