Robustness Gym: Unifying the NLP Evaluation Landscape

Despite impressive performance on standard benchmarks, deep neural networks are often brittle when deployed in real-world systems. Consequently, recent research has focused on testing the robustness of such models, resulting in a diverse set of evaluation methodologies ranging from adversarial attacks to rule-based data transformations. In this work, we identify challenges with evaluating NLP systems and propose a solution in the form of Robustness Gym (RG), a simple and extensible evaluation toolkit that unifies 4 standard evaluation paradigms: subpopulations, transformations, evaluation sets, and adversarial attacks. By providing a common platform for evaluation, Robustness Gym enables practitioners to compare results from all 4 evaluation paradigms with just a few clicks, and to easily develop and share novel evaluation methods using a built-in set of abstractions. To validate Robustness Gym's utility to practitioners, we conducted a real-world case study with a sentiment-modeling team, revealing performance degradations of 18%+. To verify that Robustness Gym can aid novel research analyses, we perform the first study of state-of-the-art commercial and academic named entity linking (NEL) systems, as well as a fine-grained analysis of state-of-the-art summarization models. For NEL, commercial systems struggle to link rare entities and lag their academic counterparts by 10%+, while state-of-the-art summarization models struggle on examples that require abstraction and distillation, degrading by 9%+. Robustness Gym can be found at https://robustnessgym.com/

FastIF: Scalable Influence Functions for Efficient Model Interpretation and Debugging

Influence functions approximate the 'influences' of training data-points for test predictions and have a wide variety of applications. Despite the popularity, their computational cost does not scale well with model and training data size. We present FastIF, a set of simple modifications to influence functions that significantly improves their run-time. We use k-Nearest Neighbors (kNN) to narrow the search space down to a subset of good candidate data points, identify the configurations that best balance the speed-quality trade-off in estimating the inverse Hessian-vector product, and introduce a fast parallel variant. Our proposed method achieves about 80x speedup while being highly correlated with the original influence values. With the availability of the fast influence functions, we demonstrate their usefulness in four applications. First, we examine whether influential data-points can 'explain' test time behavior using the framework of simulatability. Second, we visualize the influence interactions between training and test data-points. Third, we show that we can correct model errors by additional fine-tuning on certain influential data-points, improving the accuracy of a trained MNLI model by 2.6% on the HANS challenge set using a small number of gradient updates. Finally, we experiment with a data-augmentation setup where we use influence functions to search for new data-points unseen during training to improve model performance. Overall, our fast influence functions can be efficiently applied to large models and datasets, and our experiments demonstrate the potential of influence functions in model interpretation and correcting model errors. Code is available at https://github.com/salesforce/fast-influence-functions

I like fish, especially dolphins: Addressing Contradictions in Dialogue Modeling

To quantify how well natural language understanding models can capture consistency in a general conversation, we introduce the DialoguE COntradiction DEtection task (DECODE) and a new conversational dataset containing both human-human and human-bot contradictory dialogues. We then compare a structured utterance-based approach of using pre-trained Transformer models for contradiction detection with the typical unstructured approach. Results reveal that: (i) our newly collected dataset is notably more effective at providing supervision for the dialogue contradiction detection task than existing NLI data including those aimed to cover the dialogue domain; (ii) the structured utterance-based approach is more robust and transferable on both analysis and out-of-distribution dialogues than its unstructured counterpart. We also show that our best contradiction detection model correlates well with human judgments and further provide evidence for its usage in both automatically evaluating and improving the consistency of state-of-the-art generative chatbots.

I like fish, especially dolphins: Addressing Contradictions in Dialogue Modelling

To quantify how well natural language understanding models can capture consistency in a general conversation, we introduce the DialoguE COntradiction DEtection task (DECODE) and a new conversational dataset containing both human-human and human-bot contradictory dialogues. We then compare a structured utterance-based approach of using pre-trained Transformer models for contradiction detection with the typical unstructured approach. Results reveal that: (i) our newly collected dataset is notably more effective at providing supervision for the dialogue contradiction detection task than existing NLI data including those aimed to cover the dialogue domain; (ii) the structured utterance-based approach is more robust and transferable on both analysis and out-of-distribution dialogues than its unstructured counterpart. We also show that our best contradiction detection model correlates well with human judgments and further provide evidence for its usage in both automatically evaluating and improving the consistency of state-of-the-art generative chatbots.

To what extent do human explanations of model behavior align with actual model behavior?

Given the increasingly prominent role NLP models (will) play in our lives, it is important to evaluate models on their alignment with human expectations of how models behave. Using Natural Language Inference (NLI) as a case study, we investigated the extent to which human-generated explanations of models' inference decisions align with how models actually make these decisions. More specifically, we defined two alignment metrics that quantify how well natural language human explanations align with model sensitivity to input words, as measured by integrated gradients. Then, we evaluated six different transformer models (the base and large versions of BERT, RoBERTa and ELECTRA), and found that the BERT-base model has the highest alignment with human-generated explanations, for both alignment metrics. Additionally, the base versions of the models we surveyed tended to have higher alignment with human-generated explanations than their larger counterparts, suggesting that increasing the number model parameters could result in worse alignment with human explanations. Finally, we find that a model's alignment with human explanations is not predicted by the model's accuracy on NLI, suggesting that accuracy and alignment are orthogonal, and both are important ways to evaluate models.

HoVer: A Dataset for Many-Hop Fact Extraction And Claim Verification

We introduce HoVer (HOppy VERification), a dataset for many-hop evidence extraction and fact verification. It challenges models to extract facts from several Wikipedia articles that are relevant to a claim and classify whether the claim is Supported or Not-Supported by the facts. In HoVer, the claims require evidence to be extracted from as many as four English Wikipedia articles and embody reasoning graphs of diverse shapes. Moreover, most of the 3/4-hop claims are written in multiple sentences, which adds to the complexity of understanding long-range dependency relations such as coreference. We show that the performance of an existing state-of-the-art semantic-matching model degrades significantly on our dataset as the number of reasoning hops increases, hence demonstrating the necessity of many-hop reasoning to achieve strong results. We hope that the introduction of this challenging dataset and the accompanying evaluation task will encourage research in many-hop fact retrieval and information verification. We make the HoVer dataset publicly available at https://hover-nlp.github.io

ArraMon: A Joint Navigation-Assembly Instruction Interpretation Task in Dynamic Environments

For embodied agents, navigation is an important ability but not an isolated goal. Agents are also expected to perform specific tasks after reaching the target location, such as picking up objects and assembling them into a particular arrangement. We combine Vision-and-Language Navigation, assembling of collected objects, and object referring expression comprehension, to create a novel joint navigation-and-assembly task, named ArraMon. During this task, the agent (similar to a PokeMON GO player) is asked to find and collect different target objects one-by-one by navigating based on natural language instructions in a complex, realistic outdoor environment, but then also ARRAnge the collected objects part-by-part in an egocentric grid-layout environment. To support this task, we implement a 3D dynamic environment simulator and collect a dataset (in English; and also extended to Hindi) with human-written navigation and assembling instructions, and the corresponding ground truth trajectories. We also filter the collected instructions via a verification stage, leading to a total of 7.7K task instances (30.8K instructions and paths). We present results for several baseline models (integrated and biased) and metrics (nDTW, CTC, rPOD, and PTC), and the large model-human performance gap demonstrates that our task is challenging and presents a wide scope for future work. Our dataset, simulator, and code are publicly available at: https://arramonunc.github.io

DORB: Dynamically Optimizing Multiple Rewards with Bandits

Policy gradients-based reinforcement learning has proven to be a promising approach for directly optimizing non-differentiable evaluation metrics for language generation tasks. However, optimizing for a specific metric reward leads to improvements in mostly that metric only, suggesting that the model is gaming the formulation of that metric in a particular way without often achieving real qualitative improvements. Hence, it is more beneficial to make the model optimize multiple diverse metric rewards jointly. While appealing, this is challenging because one needs to manually decide the importance and scaling weights of these metric rewards. Further, it is important to consider using a dynamic combination and curriculum of metric rewards that flexibly changes over time. Considering the above aspects, in our work, we automate the optimization of multiple metric rewards simultaneously via a multi-armed bandit approach (DORB), where at each round, the bandit chooses which metric reward to optimize next, based on expected arm gains. We use the Exp3 algorithm for bandits and formulate two approaches for bandit rewards: (1) Single Multi-reward Bandit (SM-Bandit); (2) Hierarchical Multi-reward Bandit (HM-Bandit). We empirically show the effectiveness of our approaches via various automatic metrics and human evaluation on two important NLG tasks: question generation and data-to-text generation, including on an unseen-test transfer setup. Finally, we present interpretable analyses of the learned bandit curriculum over the optimized rewards.

ConjNLI: Natural Language Inference Over Conjunctive Sentences

Reasoning about conjuncts in conjunctive sentences is important for a deeper understanding of conjunctions in English and also how their usages and semantics differ from conjunctive and disjunctive boolean logic. Existing NLI stress tests do not consider non-boolean usages of conjunctions and use templates for testing such model knowledge. Hence, we introduce ConjNLI, a challenge stress-test for natural language inference over conjunctive sentences, where the premise differs from the hypothesis by conjuncts removed, added, or replaced. These sentences contain single and multiple instances of coordinating conjunctions ("and", "or", "but", "nor") with quantifiers, negations, and requiring diverse boolean and non-boolean inferences over conjuncts. We find that large-scale pre-trained language models like RoBERTa do not understand conjunctive semantics well and resort to shallow heuristics to make inferences over such sentences. As some initial solutions, we first present an iterative adversarial fine-tuning method that uses synthetically created training data based on boolean and non-boolean heuristics. We also propose a direct model advancement by making RoBERTa aware of predicate semantic roles. While we observe some performance gains, ConjNLI is still challenging for current methods, thus encouraging interesting future work for better understanding of conjunctions. Our data and code are publicly available at: https://github.com/swarnaHub/ConjNLI