EMNLP 2022 论文列表

A long-running goal of the clinical NLP community is the extraction of important variables trapped in clinical notes. However, roadblocks have included dataset shift from the general domain and a lack of public clinical corpora and annotations. In this work, we show that large language models, such as InstructGPT (Ouyang et al., 2022), perform well at zero- and few-shot information extraction from clinical text despite not being trained specifically for the clinical domain. Whereas text classification and generation performance have already been studied extensively in such models, here we additionally demonstrate how to leverage them to tackle a diverse set of NLP tasks which require more structured outputs, including span identification, token-level sequence classification, and relation extraction. Further, due to the dearth of available data to evaluate these systems, we introduce new datasets for benchmarking few-shot clinical information extraction based on a manual re-annotation of the CASI dataset (Moon et al., 2014) for new tasks. On the clinical extraction tasks we studied, the GPT-3 systems significantly outperform existing zero- and few-shot baselines.

It has been shown that dual encoders trained on one domain often fail to generalize to other domains for retrieval tasks. One widespread belief is that the bottleneck layer of a dual encoder, where the final score is simply a dot-product between a query vector and a passage vector, is too limited compared to models with fine-grained interactions between the query and the passage. In this paper, we challenge this belief by scaling up the size of the dual encoder model while keeping the bottleneck layer as a single dot-product with a fixed size. With multi-stage training, scaling up the model size brings significant improvement on a variety of retrieval tasks, especially for out-of-domain generalization. We further analyze the impact of the bottleneck layer and demonstrate diminishing improvement when scaling up the embedding size. Experimental results show that our dual encoders, Generalizable T5-based dense Retrievers (GTR), outperform previous sparse and dense retrievers on the BEIR dataset significantly. Most surprisingly, our ablation study finds that GTR is very data efficient, as it only needs 10% of MS Marco supervised data to match the out-of-domain performance of using all supervised data.

The way we talk carries information about who we are. Demographics, personality, clinical conditions, political preferences influence what we speak about and how, suggesting that many individual attributes could be inferred from adequate encodings of linguistic behavior. Conversely, conditioning text representations on author attributes has been shown to improve model performance in many NLP tasks. Previous research on individual differences and language representations has mainly focused on predicting selected attributes from text, or on conditioning text representations on such attributes for author-based contextualization. Here, we present a self-supervised approach to learning language-based user encodings using transformers. Using a large corpus of Reddit submissions, we fine-tune DistilBERT on user-based triplet loss. We show that fine-tuned models can pick up on complex linguistic signatures of users, and that they are able to infer rich information about them. Through a series of intrinsic analyses and probing tasks, we provide evidence that fine-tuning enhances models’ ability to abstract generalizable user information, which yields performance advantages for user-based downstream tasks. We discuss applications in language-based assessment and contextualized and personalized NLP.

Visual Question Answering (VQA) models are prone to learn the shortcut solution formed by dataset biases rather than the intended solution. To evaluate the VQA models’ reasoning ability beyond shortcut learning, the VQA-CP v2 dataset introduces a distribution shift between the training and test set given a question type. In this way, the model cannot use the training set shortcut (from question type to answer) to perform well on the test set. However, VQA-CP v2 only considers one type of shortcut and thus still cannot guarantee that the model relies on the intended solution rather than a solution specific to this shortcut. To overcome this limitation, we propose a new dataset that considers varying types of shortcuts by constructing different distribution shifts in multiple OOD test sets. In addition, we overcome the three troubling practices in the use of VQA-CP v2, e.g., selecting models using OOD test sets, and further standardize OOD evaluation procedure. Our benchmark provides a more rigorous and comprehensive testbed for shortcut learning in VQA. We benchmark recent methods and find that methods specifically designed for particular shortcuts fail to simultaneously generalize to our varying OOD test sets. We also systematically study the varying shortcuts and provide several valuable findings, which may promote the exploration of shortcut learning in VQA.

Language models (LMs) have recently been shown to generate more factual responses by employing modularity (Zhou et al., 2022) in combination with retrieval (Adolphs et al., 2021). We extend the recent approach of Adolphs et al. (2021) to include internet search as a module. Our SeeKeR (Search engine->Knowledge->Response) method thus applies a single LM to three modular tasks in succession: search, generating knowledge, and generating a final response. We show that, when using SeeKeR as a dialogue model, it outperforms the state-of-the-art model BlenderBot 2 (Chen et al., 2021) on open-domain knowledge-grounded conversations for the same number of parameters, in terms of consistency, knowledge and per-turn engagingness. SeeKeR applied to topical prompt completions as a standard language model outperforms GPT2 (Radford et al., 2019) and GPT3 (Brown et al., 2020) in terms of factuality and topicality, despite GPT3 being a vastly larger model. Our code and models are made publicly available.

We address the general task of structured commonsense reasoning: given a natural language input, the goal is to generate a graph such as an event or a reasoning-graph.To employ large language models (LMs) for this task, existing approaches ‘serialize’ the output graph as a flat list of nodes and edges.Although feasible, these serialized graphs strongly deviate from the natural language corpora that LMs were pre-trained on, hindering LMs from generating them correctly. In this paper, we show that when we instead frame structured commonsense reasoning tasks as code generation tasks, pre-trained LMs of code are better structured commonsense reasoners than LMs of natural language, even when the downstream task does not involve source code at all.We demonstrate our approach across three diverse structured commonsense reasoning tasks. In all these natural language tasks, we show that using our approach, a code generation LM (codex) outperforms natural-LMs that are fine-tuned on the target task (T5) and other strong LMs such as GPT-3 in the few-shot setting.

Language models (LMs) are trained on collections of documents, written by individual human agents to achieve specific goals in the outside world. During training, LMs have access only to text of these documents, with no direct evidence of the internal states of the agents that produced them—a fact often used to argue that LMs are incapable of modeling goal-directed aspects of human language production and comprehension. Can LMs trained on text learn anything at all about the relationship between language and use? I argue that LMs are models of communicative intentions in a specific, narrow sense. When performing next word prediction given a textual context, an LM can infer and represent properties of an agent likely to have produced that context. These representations can in turn influence subsequent LM generation in the same way that agents’ communicative intentions influence their language. I survey findings from the recent literature showing that—even in today’s non-robust and error-prone models—LMs infer and use representations of fine-grained communicative intentions and high-level beliefs and goals. Despite the limited nature of their training data, they can thus serve as building blocks for systems that communicate and act intentionally.

Some claim language models understand us. Others won’t hear it. To clarify, I investigate three views of human language understanding: as-mapping, as-reliability and as-representation. I argue that while behavioral reliability is necessary for understanding, internal representations are sufficient; they climb the right hill. I review state-of-the-art language and multi-modal models: they are pragmatically challenged by under-specification of form. I question the Scaling Paradigm: limits on resources may prohibit scaled-up models from approaching understanding. Last, I describe how as-representation advances a science of understanding. We need work which probes model internals, adds more of human language, and measures what models can learn.

We examine LMs’ competence of directional predicate entailments by supervised fine-tuning with prompts. Our analysis shows that contrary to their apparent success on standard NLI, LMs show limited ability to learn such directional inference; moreover, existing datasets fail to test directionality, and/or are infested by artefacts that can be learnt as proxy for entailments, yielding over-optimistic results. In response, we present BoOQA (Boolean Open QA), a robust multi-lingual evaluation benchmark for directional predicate entailments, extrinsic to existing training sets. On BoOQA, we establish baselines and show evidence of existing LM-prompting models being incompetent directional entailment learners, in contrast to entailment graphs, however limited by sparsity.

Modern large-scale Pre-trained Language Models (PLMs) have achieved tremendous success on a wide range of downstream tasks. However, most of the LM pre-training objectives only focus on text reconstruction, but have not sought to learn latent-level interpretable representations of sentences. In this paper, we manage to push the language models to obtain a deeper understanding of sentences by proposing a new pre-training objective, Sparse Latent Typing, which enables the model to sparsely extract sentence-level keywords with diverse latent types. Experimental results show that our model is able to learn interpretable latent type categories in a self-supervised manner without using any external knowledge. Besides, the language model pre-trained with such an objective also significantly improves Information Extraction related downstream tasks in both supervised and few-shot settings. Our code is publicly available at https://github.com/renll/SparseLT.

To reduce the toxic degeneration in a pretrained Language Model (LM), previous work on Language Model detoxification has focused on reducing the toxicity of the generation itself (self-toxicity) without consideration of the context. As a result, a type of implicit offensive language where the generations support the offensive language in the context is ignored. Different from the LM controlling tasks in previous work, where the desired attributes are fixed for generation, the desired stance of the generation depends on the offensiveness of the context. Therefore, we propose a novel control method to do context-dependent detoxification with the stance taken into consideration. We introduce meta prefixes to learn the contextualized stance control strategy and to generate the stance control prefix according to the input context. The generated stance prefix is then combined with the toxicity control prefix to guide the response generation. Experimental results show that our proposed method can effectively learn the context-dependent stance control strategies while keeping a low self-toxicity of the underlying LM.

Pre-trained language models (LMs), such as BERT (Devlin et al., 2018) and its variants, have led to significant improvements on various NLP tasks in past years. However, a theoretical framework for studying their relationships is still missing. In this paper, we fill this gap by investigating the linear dependency between pre-trained LMs. The linear dependency of LMs is defined analogously to the linear dependency of vectors. We propose Language Model Decomposition (LMD) to represent a LM using a linear combination of other LMs as basis, and derive the closed-form solution. A goodness-of-fit metric for LMD similar to the coefficient of determination is defined and used to measure the linear dependency of a set of LMs. In experiments, we find that BERT and eleven (11) BERT-like LMs are 91% linearly dependent. This observation suggests that current state-of-the-art (SOTA) LMs are highly “correlated”. To further advance SOTA we need more diverse and novel LMs that are less dependent on existing LMs.

English pretrained language models, which make up the backbone of many modern NLP systems, require huge amounts of unlabeled training data. These models are generally presented as being trained only on English text but have been found to transfer surprisingly well to other languages. We investigate this phenomenon and find that common English pretraining corpora actually contain significant amounts of non-English text: even when less than 1% of data is not English (well within the error rate of strong language classifiers), this leads to hundreds of millions of foreign language tokens in large-scale datasets. We then demonstrate that even these small percentages of non-English data facilitate cross-lingual transfer for models trained on them, with target language performance strongly correlated to the amount of in-language data seen during pretraining. In light of these findings, we argue that no model is truly monolingual when pretrained at scale, which should be considered when evaluating cross-lingual transfer.

Despite the great success of spoken language understanding (SLU) in high-resource languages, it remains challenging in low-resource languages mainly due to the lack of labeled training data. The recent multilingual code-switching approach achieves better alignments of model representations across languages by constructing a mixed-language context in zero-shot cross-lingual SLU. However, current code-switching methods are limited to implicit alignment and disregard the inherent semantic structure in SLU, i.e., the hierarchical inclusion of utterances, slots and words. In this paper, we propose to model the utterance-slot-word structure by a multi-level contrastive learning framework at the utterance, slot and word levels to facilitate explicit alignment. Novel code-switching schemes are introduced to generate hard negative examples for our contrastive learning framework. Furthermore, we develop a label-aware joint model leveraging label semantics to enhance the implicit alignment and feed to contrastive learning. Our experimental results show that our proposed methods significantly improve the performance compared with the strong baselines on two zero-shot cross-lingual SLU benchmark datasets.

Multi-Label Few-Shot Aspect Category Detection (FS-ACD) is a new sub-task of aspect-based sentiment analysis, which aims to detect aspect categories accurately with limited training instances. Recently, dominant works use the prototypical network to accomplish this task, and employ the attention mechanism to extract keywords of aspect category from the sentences to produce the prototype for each aspect. However, they still suffer from serious noise problems: (1) due to lack of sufficient supervised data, the previous methods easily catch noisy words irrelevant to the current aspect category, which largely affects the quality of the generated prototype; (2) the semantically-close aspect categories usually generate similar prototypes, which are mutually noisy and confuse the classifier seriously. In this paper, we resort to the label information of each aspect to tackle the above problems, along with proposing a novel Label-Driven Denoising Framework (LDF). Extensive experimental results show that our framework achieves better performance than other state-of-the-art methods.

Multimodal Machine Translation (MMT) focuses on enhancing text-only translation with visual features, which has attracted considerable attention from both natural language processing and computer vision communities. Recent advances still struggle to train a separate model for each language pair, which is costly and unaffordable when the number of languages increases in the real world. In other words, the multilingual multimodal machine translation (Multilingual MMT) task has not been investigated, which aims to handle the aforementioned issues by providing a shared semantic space for multiple languages. Besides, the image modality has no language boundaries, which is superior to bridging the semantic gap between languages. To this end,we first propose the Multilingual MMT task by establishing two new Multilingual MMT benchmark datasets covering seven languages.Then, an effective baseline LVP-M3 using visual prompts is proposed to support translations between different languages,which includes three stages (token encoding, language-aware visual prompt generation, and language translation). Extensive experimental results on our constructed benchmark datasets demonstrate the effectiveness of LVP-M3 method for Multilingual MMT.

When we execute the typical fine-tuning paradigm on continuously sequential tasks, the model will suffer from the catastrophic forgetting problem (i.e., the model tends to adjust old parameters according to the new knowledge, which leads to the loss of previously acquired concepts). People proposed replay-based methods by accessing old data from extra storage and maintaining the parameters of old concepts, which actually raise the privacy issue and larger memory requirements. In this work, we aim to achieve the sequential/continual learning of knowledge without accessing the old data. The core idea is to calibrate the parameters and logits (output) so that preserving old parameters and generalized learning on new concepts can be solved simultaneously. Our proposed framework includes two major components, Logits Calibration (LC) and Parameter Calibration (PC). The LC focuses on calibrating the learning of novel models with old models, and PC aims to preserve the parameters of old models. These two operations can maintain the old knowledge while learning new tasks without storing previous data. We conduct experiments on various scenarios of the GLUE (the General Language Understanding Evaluation) benchmark. The experimental results show that our model achieves state-of-the-art performance in all scenarios.

Weakly supervised text classification methods typically train a deep neural classifier based on pseudo-labels. The quality of pseudo-labels is crucial to final performance but they are inevitably noisy due to their heuristic nature, so selecting the correct ones has a huge potential for performance boost. One straightforward solution is to select samples based on the softmax probability scores in the neural classifier corresponding to their pseudo-labels. However, we show through our experiments that such solutions are ineffective and unstable due to the erroneously high-confidence predictions from poorly calibrated models. Recent studies on the memorization effects of deep neural models suggest that these models first memorize training samples with clean labels and then those with noisy labels. Inspired by this observation, we propose a novel pseudo-label selection method LOPS that takes learning order of samples into consideration. We hypothesize that the learning order reflects the probability of wrong annotation in terms of ranking, and therefore, propose to select the samples that are learnt earlier. LOPS can be viewed as a strong performance-boost plug-in to most existing weakly-supervised text classification methods, as confirmed in extensive experiments on four real-world datasets.

Mathematical reasoning skills are essential for general-purpose intelligentsystems to perform tasks from grocery shopping to climate modeling.Towards evaluating and improving AI systems in this domain, we proposeLILA, a unified mathematical reasoning benchmark consisting of 23 diversetasks along four dimensions:(i) mathematical abilities e.g., arithmetic, calculus (ii) language format e.g., question-answering, fill-in-the-blanks (iii) language diversity e.g., no language, simple language (iv) external knowledge e.g., commonsense, physics. We construct our benchmark by extending 20 datasets benchmark by collecting task instructions and solutions in the form of Python programs,thereby obtaining explainable solutions in addition to the correct answer.We additionally introduce two evaluation datasets to measure out-of-distribution performance and robustness to language perturbation.Finally, we introduce BHASKARA,a general-purpose mathematical reasoning model trained on LILA. Importantly, we find that multi-tasking leads to significant improvements (average relative improvement of 21.83% F1 score vs. single-task models),while the best performing model only obtains 60.40%,indicating the room for improvement in general mathematical reasoning and understanding.