Typical fact verification models use retrieved written evidence to verify claims. Evidence sources, however, often change over time as more information is gathered and revised. In order to adapt, models must be sensitive to subtle differences in supporting evidence. We present VitaminC, a benchmark infused with challenging cases that require fact verification models to discern and adjust to slight factual changes. We collect over 100,000 Wikipedia revisions that modify an underlying fact, and leverage these revisions, together with additional synthetically constructed ones, to create a total of over 400,000 claim-evidence pairs. Unlike previous resources, the examples in VitaminC are contrastive, i.e., they contain evidence pairs that are nearly identical in language and content, with the exception that one supports a given claim while the other does not. We show that training using this design increases robustness -- improving accuracy by 10% on adversarial fact verification and 6% on adversarial natural language inference (NLI). Moreover, the structure of VitaminC leads us to define additional tasks for fact-checking resources: tagging relevant words in the evidence for verifying the claim, identifying factual revisions, and providing automatic edits via factually consistent text generation.
In meta-learning, the knowledge learned from previous tasks is transferred to new ones, but this transfer only works if tasks are related. Sharing information between unrelated tasks might hurt performance, and it is unclear how to transfer knowledge across tasks with a hierarchical structure. Our research extends a model agnostic meta-learning model, MAML, by exploiting hierarchical task relationships. Our algorithm, TreeMAML, adapts the model to each task with a few gradient steps, but the adaptation follows the hierarchical tree structure: in each step, gradients are pooled across tasks clusters, and subsequent steps follow down the tree. We also implement a clustering algorithm that generates the tasks tree without previous knowledge of the task structure, allowing us to make use of implicit relationships between the tasks. We show that the new algorithm, which we term TreeMAML, performs better than MAML when the task structure is hierarchical for synthetic experiments. To study the performance of the method in real-world data, we apply this method to Natural Language Understanding, we use our algorithm to finetune Language Models taking advantage of the language phylogenetic tree. We show that TreeMAML improves the state of the art results for cross-lingual Natural Language Inference. This result is useful, since most languages in the world are under-resourced and the improvement on cross-lingual transfer allows the internationalization of NLP models. This results open the window to use this algorithm in other real-world hierarchical datasets.
Modern natural language understanding models depend on pretrained subword embeddings, but applications may need to reason about words that were never or rarely seen during pretraining. We show that examples that depend critically on a rarer word are more challenging for natural language inference models. Then we explore how a model could learn to use definitions, provided in natural text, to overcome this handicap. Our model's understanding of a definition is usually weaker than a well-modeled word embedding, but it recovers most of the performance gap from using a completely untrained word.
Natural Language Processing algorithms have made incredible progress, but they still struggle when applied to out-of-distribution examples. We address a challenging and underexplored version of this domain adaptation problem, where an algorithm is trained on several source domains, and then applied to examples from unseen domains that are unknown at training time. Particularly, no examples, labeled or unlabeled, or any other knowledge about the target domain are available to the algorithm at training time. We present PADA: An example-based autoregressive Prompt learning algorithm for on-the-fly Any-Domain Adaptation, based on the T5 language model. Given a test example, PADA first generates a unique prompt for it and then, conditioned on this prompt, labels the example with respect to the NLP prediction task. PADA is trained to generate a prompt which is a token sequence of unrestricted length, consisting of Domain Related Features (DRFs) that characterize each of the source domains. Intuitively, the generated prompt is a unique signature that maps the test example to a semantic space spanned by the source domains. In experiments with 3 tasks (text classification and sequence tagging), for a total of 14 multi-source adaptation scenarios, PADA substantially outperforms strong baselines.
Recent works have demonstrated reasonable success of representation learning in hypercomplex space. Specifically, "fully-connected layers with Quaternions" (4D hypercomplex numbers), which replace real-valued matrix multiplications in fully-connected layers with Hamilton products of Quaternions, both enjoy parameter savings with only 1/4 learnable parameters and achieve comparable performance in various applications. However, one key caveat is that hypercomplex space only exists at very few predefined dimensions (4D, 8D, and 16D). This restricts the flexibility of models that leverage hypercomplex multiplications. To this end, we propose parameterizing hypercomplex multiplications, allowing models to learn multiplication rules from data regardless of whether such rules are predefined. As a result, our method not only subsumes the Hamilton product, but also learns to operate on any arbitrary nD hypercomplex space, providing more architectural flexibility using arbitrarily $1/n$ learnable parameters compared with the fully-connected layer counterpart. Experiments of applications to the LSTM and Transformer models on natural language inference, machine translation, text style transfer, and subject verb agreement demonstrate architectural flexibility and effectiveness of the proposed approach.
With the success of pre-trained language models in recent years, more and more researchers focus on opening the "black box" of these models. Following this interest, we carry out a qualitative and quantitative analysis of constituency grammar in attention heads of BERT and RoBERTa. We employ the syntactic distance method to extract implicit constituency grammar from the attention weights of each head. Our results show that there exist heads that can induce some grammar types much better than baselines, suggesting that some heads act as a proxy for constituency grammar. We also analyze how attention heads' constituency grammar inducing (CGI) ability changes after fine-tuning with two kinds of tasks, including sentence meaning similarity (SMS) tasks and natural language inference (NLI) tasks. Our results suggest that SMS tasks decrease the average CGI ability of upper layers, while NLI tasks increase it. Lastly, we investigate the connections between CGI ability and natural language understanding ability on QQP and MNLI tasks.
We study estimating inherent human disagreement (annotation label distribution) in natural language inference task. Post-hoc smoothing of the predicted label distribution to match the expected label entropy is very effective. Such simple manipulation can reduce KL divergence by almost half, yet will not improve majority label prediction accuracy or learn label distributions. To this end, we introduce a small amount of examples with multiple references into training. We depart from the standard practice of collecting a single reference per each training example, and find that collecting multiple references can achieve better accuracy under the fixed annotation budget. Lastly, we provide rich analyses comparing these two methods for improving label distribution estimation.
Lexical inference in context (LIiC) is the task of recognizing textual entailment between two very similar sentences, i.e., sentences that only differ in one expression. It can therefore be seen as a variant of the natural language inference task that is focused on lexical semantics. We formulate and evaluate the first approaches based on pretrained language models (LMs) for this task: (i) a few-shot NLI classifier, (ii) a relation induction approach based on handcrafted patterns expressing the semantics of lexical inference, and (iii) a variant of (ii) with patterns that were automatically extracted from a corpus. All our approaches outperform the previous state of the art, showing the potential of pretrained LMs for LIiC. In an extensive analysis, we investigate factors of success and failure of our three approaches.
Transformers have emerged as a powerful tool for a broad range of natural language processing tasks. A key component that drives the impressive performance of Transformers is the self-attention mechanism that encodes the influence or dependence of other tokens on each specific token. While beneficial, the quadratic complexity of self-attention on the input sequence length has limited its application to longer sequences -- a topic being actively studied in the community. To address this limitation, we propose Nystr\"{o}mformer -- a model that exhibits favorable scalability as a function of sequence length. Our idea is based on adapting the Nystr\"{o}m method to approximate standard self-attention with $O(n)$ complexity. The scalability of Nystr\"{o}mformer enables application to longer sequences with thousands of tokens. We perform evaluations on multiple downstream tasks on the GLUE benchmark and IMDB reviews with standard sequence length, and find that our Nystr\"{o}mformer performs comparably, or in a few cases, even slightly better, than standard self-attention. On longer sequence tasks in the Long Range Arena (LRA) benchmark, Nystr\"{o}mformer performs favorably relative to other efficient self-attention methods. Our code is available at https://github.com/mlpen/Nystromformer.
In the last couple of years there were a few attempts to apply topological data analysis to text, and in particular to natural language inference. A recent work by Tymochko et al. suggests the possibility of capturing `the notion of logical shape in text,' using `topological delay embeddings,' a technique derived from dynamical systems, applied to word embeddings. In this note we reconstruct their argument and show, using several old and new examples, that the problem of connecting logic, topology and text is still very much unsolved. We conclude that there is no clear answer to the question: ``Can we find a circle in a circular argument?'' We point out some possible avenues of exploration. The code used in our experiment is also shown.
Fine-grained visual recognition, which aims to identify subcategories of the same base-level category, is a challenging task because of its large intra-class variances and small inter-class variances. Human beings can perform object recognition task based on not only the visual appearance but also the knowledge from texts, as texts can point out the discriminative parts or characteristics which are always the key to distinguishing different subcategories. This is an involuntary transfer from human textual attention to visual attention, suggesting that texts are able to assist fine-grained recognition. In this paper, we propose a Text-Embedded Bilinear (TEB) model which incorporates texts as extra guidance for fine-grained recognition. Specially, we first conduct a text-embedded network to embed text feature into the discriminative image feature learning to get a embedded feature. In addition, since the cross-layer part feature interaction and fine-grained feature learning are mutually correlated and can reinforce each other, we also extract a candidate feature from the text encoder and embed it into the inter-layer feature of the image encoder to get an embedded candidate feature. At last we utilize a cross-layer bilinear network to fuse the two embedded features. Comparing with state-of-the-art methods on the widely used CUB-200-2011 dataset and Oxford Flowers-102 dataset for fine-grained image recognition, the experimental results demonstrate our TEB model achieves the best performance.
Open Information Extraction (OIE) systems seek to compress the factual propositions of a sentence into a series of n-ary tuples. These tuples are useful for downstream tasks in natural language processing like knowledge base creation, textual entailment, and natural language understanding. However, current OIE datasets are limited in both size and diversity. We introduce a new dataset by converting the QA-SRL 2.0 dataset to a large-scale OIE dataset (LSOIE). Our LSOIE dataset is 20 times larger than the next largest human-annotated OIE dataset. We construct and evaluate several benchmark OIE models on LSOIE, providing baselines for future improvements on the task. Our LSOIE data, models, and code are made publicly available
We propose pre-finetuning, an additional large-scale learning stage between language model pre-training and fine-tuning. Pre-finetuning is massively multi-task learning (around 50 datasets, over 4.8 million total labeled examples), and is designed to encourage learning of representations that generalize better to many different tasks. We show that pre-finetuning consistently improves performance for pretrained discriminators (e.g.~RoBERTa) and generation models (e.g.~BART) on a wide range of tasks (sentence prediction, commonsense reasoning, MRC, etc.), while also significantly improving sample efficiency during fine-tuning. We also show that large-scale multi-tasking is crucial; pre-finetuning can hurt performance when few tasks are used up until a critical point (usually above 15) after which performance improves linearly in the number of tasks.
Despite the recent success of deep neural networks in natural language processing, the extent to which they can demonstrate human-like generalization capacities for natural language understanding remains unclear. We explore this issue in the domain of natural language inference (NLI), focusing on the transitivity of inference relations, a fundamental property for systematically drawing inferences. A model capturing transitivity can compose basic inference patterns and draw new inferences. We introduce an analysis method using synthetic and naturalistic NLI datasets involving clause-embedding verbs to evaluate whether models can perform transitivity inferences composed of veridical inferences and arbitrary inference types. We find that current NLI models do not perform consistently well on transitivity inference tasks, suggesting that they lack the generalization capacity for drawing composite inferences from provided training examples. The data and code for our analysis are publicly available at https://github.com/verypluming/transitivity.
Contextualized representations from a pre-trained language model are central to achieve a high performance on downstream NLP task. The pre-trained BERT and A Lite BERT (ALBERT) models can be fine-tuned to give state-ofthe-art results in sentence-pair regressions such as semantic textual similarity (STS) and natural language inference (NLI). Although BERT-based models yield the [CLS] token vector as a reasonable sentence embedding, the search for an optimal sentence embedding scheme remains an active research area in computational linguistics. This paper explores on sentence embedding models for BERT and ALBERT. In particular, we take a modified BERT network with siamese and triplet network structures called Sentence-BERT (SBERT) and replace BERT with ALBERT to create Sentence-ALBERT (SALBERT). We also experiment with an outer CNN sentence-embedding network for SBERT and SALBERT. We evaluate performances of all sentence-embedding models considered using the STS and NLI datasets. The empirical results indicate that our CNN architecture improves ALBERT models substantially more than BERT models for STS benchmark. Despite significantly fewer model parameters, ALBERT sentence embedding is highly competitive to BERT in downstream NLP evaluations.
Natural Language Inference (NLI) or Recognizing Textual Entailment (RTE) is the task of predicting the entailment relation between a pair of sentences (premise and hypothesis). This task has been described as a valuable testing ground for the development of semantic representations, and is a key component in natural language understanding evaluation benchmarks. Models that understand entailment should encode both, the premise and the hypothesis. However, experiments by Poliak et al. revealed a strong preference of these models towards patterns observed only in the hypothesis, based on a 10 dataset comparison. Their results indicated the existence of statistical irregularities present in the hypothesis that bias the model into performing competitively with the state of the art. While recast datasets provide large scale generation of NLI instances due to minimal human intervention, the papers that generate them do not provide fine-grained analysis of the potential statistical patterns that can bias NLI models. In this work, we analyze hypothesis-only models trained on one of the recast datasets provided in Poliak et al. for word-level patterns. Our results indicate the existence of potential lexical biases that could contribute to inflating the model performance.
Dialog systems enriched with external knowledge can handle user queries that are outside the scope of the supporting databases/APIs. In this paper, we follow the baseline provided in DSTC9 Track 1 and propose three subsystems, KDEAK, KnowleDgEFactor, and Ens-GPT, which form the pipeline for a task-oriented dialog system capable of accessing unstructured knowledge. Specifically, KDEAK performs knowledge-seeking turn detection by formulating the problem as natural language inference using knowledge from dialogs, databases and FAQs. KnowleDgEFactor accomplishes the knowledge selection task by formulating a factorized knowledge/document retrieval problem with three modules performing domain, entity and knowledge level analyses. Ens-GPT generates a response by first processing multiple knowledge snippets, followed by an ensemble algorithm that decides if the response should be solely derived from a GPT2-XL model, or regenerated in combination with the top-ranking knowledge snippet. Experimental results demonstrate that the proposed pipeline system outperforms the baseline and generates high-quality responses, achieving at least 58.77% improvement on BLEU-4 score.
We present SICK-NL (read: signal), a dataset targeting Natural Language Inference in Dutch. SICK-NL is obtained by translating the SICK dataset of Marelli et al. (2014)from English into Dutch. Having a parallel inference dataset allows us to compare both monolingual and multilingual NLP models for English and Dutch on the two tasks. In the paper, we motivate and detail the translation process, perform a baseline evaluation on both the original SICK dataset and its Dutch incarnation SICK-NL, taking inspiration from Dutch skipgram embeddings and contextualised embedding models. In addition, we encapsulate two phenomena encountered in the translation to formulate stress tests and verify how well the Dutch models capture syntactic restructurings that do not affect semantics. Our main finding is all models perform worse on SICK-NL than on SICK, indicating that the Dutch dataset is more challenging than the English original. Results on the stress tests show that models don't fully capture word order freedom in Dutch, warranting future systematic studies.
Transformer based models, like BERT and RoBERTa, have achieved state-of-the-art results in many Natural Language Processing tasks. However, their memory footprint, inference latency, and power consumption are prohibitive efficient inference at the edge, and even at the data center. While quantization can be a viable solution for this, previous work on quantizing Transformer based models use floating-point arithmetic during inference, which cannot efficiently utilize integer-only logical units such as the recent Turing Tensor Cores, or traditional integer-only ARM processors. In this work, we propose I-BERT, a novel quantization scheme for Transformer based models that quantizes the entire inference with integer-only arithmetic. Based on lightweight integer-only approximation methods for nonlinear operations, e.g., GELU, Softmax, and Layer Normalization, I-BERT performs an end-to-end integer-only BERT inference without any floating point calculation. We evaluate our approach on GLUE downstream tasks using RoBERTa-Base/Large. We show that for both cases, I-BERT achieves similar (and slightly higher) accuracy as compared to the full-precision baseline. Furthermore, our preliminary implementation of I-BERT shows a speedup of 2.4-4.0x for INT8 inference on a T4 GPU system as compared to FP32 inference. The framework has been developed in PyTorch and has been open-sourced.
Recent investigations into the inner-workings of state-of-the-art large-scale pre-trained Transformer-based Natural Language Understanding (NLU) models indicate that they appear to know humanlike syntax, at least to some extent. We provide novel evidence that complicates this claim: we find that state-of-the-art Natural Language Inference (NLI) models assign the same labels to permuted examples as they do to the original, i.e. they are largely invariant to random word-order permutations. This behavior notably differs from that of humans; we struggle with ungrammatical sentences. To measure the severity of this issue, we propose a suite of metrics and investigate which properties of particular permutations lead models to be word-order invariant. In the MNLI dataset, for example, we find almost all (98.7%) examples contain at least one permutation which elicits the gold label. Models are sometimes even able to assign gold labels to permutations that they originally failed to predict correctly. We provide a comprehensive empirical evaluation of this phenomenon, and further show that this issue exists for both Transformers and pre-Transformer RNN / ConvNet based encoders, as well as across multiple languages (English and Mandarin Chinese). Our code and data are available at https://github.com/facebookresearch/unlu.