EMNLP 2022 论文列表

Despite of the superb performance on a wide range of tasks, pre-trained language models (e.g., BERT) have been proved vulnerable to adversarial texts. In this paper, we present RoChBERT, a framework to build more Robust BERT-based models by utilizing a more comprehensive adversarial graph to fuse Chinese phonetic and glyph features into pre-trained representations during fine-tuning. Inspired by curriculum learning, we further propose to augment the training dataset with adversarial texts in combination with intermediate samples. Extensive experiments demonstrate that RoChBERT outperforms previous methods in significant ways: (i) robust – RoChBERT greatly improves the model robustness without sacrificing accuracy on benign texts. Specifically, the defense lowers the success rates of unlimited and limited attacks by 59.43% and 39.33% respectively, while remaining accuracy of 93.30%; (ii) flexible – RoChBERT can easily extend to various language models to solve different downstream tasks with excellent performance; and (iii) efficient – RoChBERT can be directly applied to the fine-tuning stage without pre-training language model from scratch, and the proposed data augmentation method is also low-cost.

Question answering models can use rich knowledge sources — up to one hundred retrieved passages and parametric knowledge in the large-scale language model (LM). Prior work assumes information in such knowledge sources is consistent with each other, paying little attention to how models blend information stored in their LM parameters with that from retrieved evidence documents. In this paper, we simulate knowledge conflicts (i.e., where parametric knowledge suggests one answer and different passages suggest different answers) and examine model behaviors. We find retrieval performance heavily impacts which sources models rely on, and current models mostly rely on non-parametric knowledgein their best-performing settings. We discover a troubling trend that contradictions among knowledge sources affect model confidence only marginally. To address this issue, we present a new calibration study, where models are discouraged from presenting any single answer when presented with multiple conflicting answer candidates in retrieved evidences.

Text games present opportunities for natural language understanding (NLU) methods to tackle reinforcement learning (RL) challenges. However, recent work has questioned the necessity of NLU by showing random text hashes could perform decently. In this paper, we pursue a fine-grained investigation into the roles of text in the face of different RL challenges, and reconcile that semantic and non-semantic language representations could be complementary rather than contrasting. Concretely, we propose a simple scheme to extract relevant contextual information into an approximate state hash as extra input for an RNN-based text agent. Such a lightweight plug-in achieves competitive performance with state-of-the-art text agents using advanced NLU techniques such as knowledge graph and passage retrieval, suggesting non-NLU methods might suffice to tackle the challenge of partial observability. However, if we remove RNN encoders and use approximate or even ground-truth state hash alone, the model performs miserably, which confirms the importance of semantic function approximation to tackle the challenge of combinatorially large observation and action spaces. Our findings and analysis provide new insights for designing better text game task setups and agents.

Recently, knowledge-enhanced pre-trained language models (KEPLMs) improve context-aware representations via learning from structured relations in knowledge bases, and/or linguistic knowledge from syntactic or dependency analysis. Unlike English, there is a lack of high-performing open-source Chinese KEPLMs in the natural language processing (NLP) community to support various language understanding applications. In this paper, we revisit and advance the development of Chinese natural language understanding with a series of novel Chinese KEPLMs released in various parameter sizes, namely CKBERT (Chinese knowledge-enhanced BERT). Specifically, both relational and linguistic knowledge is effectively injected into CKBERT based on two novel pre-training tasks, i.e., linguistic-aware masked language modeling and contrastive multi-hop relation modeling. Based on the above two pre-training paradigms and our in-house implemented TorchAccelerator, we have pre-trained base (110M), large (345M) and huge (1.3B) versions of CKBERT efficiently on GPU clusters. Experiments demonstrate that CKBERT consistently outperforms strong baselines for Chinese over various benchmark NLP tasks and in terms of different model sizes.

The recent literature in text classification is biased towards short text sequences (e.g., sentences or paragraphs). In real-world applications, multi-page multi-paragraph documents are common and they cannot be efficiently encoded by vanilla Transformer-based models. We compare different Transformer-based Long Document Classification (TrLDC) approaches that aim to mitigate the computational overhead of vanilla transformers to encode much longer text, namely sparse attention and hierarchical encoding methods.We examine several aspects of sparse attention (e.g., size of local attention window, use of global attention) and hierarchical (e.g., document splitting strategy) transformers on four document classification datasets covering different domains. We observe a clear benefit from being able to process longer text, and, based on our results, we derive practical advice of applying Transformer-based models on long document classification tasks.

There is a growing body of work in recent years to develop pre-trained language models (PLMs) for the Arabic language. This work addresses two major problems in existing Arabic PLMs that limit the progress of the Arabic NLU and NLG fields. First, existing Arabic PLMs are not well-explored and their pre-training can be improved significantly using a more methodical approach. Second, there is a lack of systematic and reproducible evaluation of these models in the literature. We revisit both the pre-training and evaluation of Arabic PLMs. In terms of pre-training, we explore the impact of the quality of the pretraining data, the size of the model, and the incorporation of character-level information on Arabic PLM. As a result, we release three new Arabic BERT-style models ( JABER, Char-JABER, and SABER), and two T5-style models (AT5S and AT5B). In terms of evaluation, we conduct a comprehensive empirical study to systematically evaluate the performance of existing state-of-the-art models on ALUE, a leaderboard-powered benchmark for Arabic NLU tasks, and on a subset of the Arabic generative tasks. We show that our models significantly outperform existing Arabic PLMs and achieve a new state-of-the-art performance on discriminative and generative Arabic NLU and NLG tasks. Our models and source code to reproduce results will be made available upon acceptance.

Parameter-Efficient Tuning (PETuning) methods have been deemed by many as the new paradigm for using pretrained language models (PLMs). By tuning just a fraction amount of parameters comparing to full model finetuning, PETuning methods claim to have achieved performance on par with or even better than finetuning. In this work, we take a step back and re-examine these PETuning methods by conducting the first comprehensive investigation into the training and evaluation of them. We found the problematic validation and testing practice in current studies, when accompanied by the instability nature of PETuning methods, has led to unreliable conclusions. When being compared under a truly fair evaluation protocol, PETuning cannot yield consistently competitive performance while finetuning remains to be the best-performing method in medium- and high-resource settings. We delve deeper into the cause of the instability and observed that the number of trainable parameters and training iterations are two main factors: reducing trainable parameters and prolonging training iterations may lead to higher stability in PETuning methods.

Pretraining-based (PT-based) automatic evaluation metrics (e.g., BERTScore and BARTScore) have been widely used in several sentence generation tasks (e.g., machine translation and text summarization) due to their better correlation with human judgments over traditional overlap-based methods. Although PT-based methods have become the de facto standard for training grammatical error correction (GEC) systems, GEC evaluation still does not benefit from pretrained knowledge. This paper takes the first step towards understanding and improving GEC evaluation with pretraining. We first find that arbitrarily applying PT-based metrics to GEC evaluation brings unsatisfactory correlation results because of the excessive attention to inessential systems outputs (e.g., unchanged parts). To alleviate the limitation, we propose a novel GEC evaluation metric to achieve the best of both worlds, namely PT-M2 which only uses PT-based metrics to score those corrected parts. Experimental results on the CoNLL14 evaluation task show that PT-M2 significantly outperforms existing methods, achieving a new state-of-the-art result of 0.949 Pearson correlation. Further analysis reveals that PT-M2 is robust to evaluate competitive GEC systems. Source code and scripts are freely available at https://github.com/pygongnlp/PT-M2.

The DocRED dataset is one of the most popular and widely used benchmarks for document-level relation extraction (RE). It adopts a recommend-revise annotation scheme so as to have a large-scale annotated dataset. However, we find that the annotation of DocRED is incomplete, i.e., false negative samples are prevalent. We analyze the causes and effects of the overwhelming false negative problem in the DocRED dataset. To address the shortcoming, we re-annotate 4,053 documents in the DocRED dataset by adding the missed relation triples back to the original DocRED. We name our revised DocRED dataset Re-DocRED. We conduct extensive experiments with state-of-the-art neural models on both datasets, and the experimental results show that the models trained and evaluated on our Re-DocRED achieve performance improvements of around 13 F1 points. Moreover, we conduct a comprehensive analysis to identify the potential areas for further improvement.

We introduce a new method to improve existing multilingual sentence embeddings with Abstract Meaning Representation (AMR). Compared with the original textual input, AMR is a structured semantic representation that presents the core concepts and relations in a sentence explicitly and unambiguously. It also helps reduce the surface variations across different expressions and languages. Unlike most prior work that only evaluates the ability to measure semantic similarity, we present a thorough evaluation of existing multilingual sentence embeddings and our improved versions, which include a collection of five transfer tasks in different downstream applications. Experiment results show that retrofitting multilingual sentence embeddings with AMR leads to better state-of-the-art performance on both semantic textual similarity and transfer tasks.

Despite pre-training’s progress in many important NLP tasks, it remains to explore effective pre-training strategies for dense retrieval. In this paper, we propose RetroMAE, a new retrieval oriented pre-training paradigm based on Masked Auto-Encoder (MAE). RetroMAE is highlighted by three critical designs. 1) A novel MAE workflow, where the input sentence is polluted for encoder and decoder with different masks. The sentence embedding is generated from the encoder’s masked input; then, the original sentence is recovered based on the sentence embedding and the decoder’s masked input via masked language modeling. 2) Asymmetric model structure, with a full-scale BERT like transformer as encoder, and a one-layer transformer as decoder. 3) Asymmetric masking ratios, with a moderate ratio for encoder: 15 30%, and an aggressive ratio for decoder: 50 70%. Our framework is simple to realize and empirically competitive: the pre-trained models dramatically improve the SOTA performances on a wide range of dense retrieval benchmarks, like BEIR and MS MARCO. The source code and pre-trained models are made publicly available at https://github.com/staoxiao/RetroMAE so as to inspire more interesting research.

Event argument extraction has long been studied as a sequential prediction problem with extractive-based methods, tackling each argument in isolation. Although recent work proposes generation-based methods to capture cross-argument dependency, they require generating and post-processing a complicated target sequence (template). Motivated by these observations and recent pretrained language models’ capabilities of learning from demonstrations. We propose a retrieval-augmented generative QA model (R-GQA) for event argument extraction. It retrieves the most similar QA pair and augments it as prompt to the current example’s context, then decodes the arguments as answers. Our approach outperforms substantially prior methods across various settings (i.e. fully supervised, domain transfer, and fewshot learning). Finally, we propose a clustering-based sampling strategy (JointEnc) and conduct a thorough analysis of how different strategies influence the few-shot learning performances.

Systems for knowledge-intensive tasks such as open-domain question answering (QA) usually consist of two stages: efficient retrieval of relevant documents from a large corpus and detailed reading of the selected documents. This is usually done through two separate models, a retriever that encodes the query and finds nearest neighbors, and a reader based on Transformers. These two components are usually modeled separately, which necessitates a cumbersome implementation and is awkward to optimize in an end-to-end fashion. In this paper, we revisit this design and eschew the separate architecture and training in favor of a single Transformer that performs retrieval as attention (RAA), and end-to-end training solely based on supervision from the end QA task. We demonstrate for the first time that an end-to-end trained single Transformer can achieve both competitive retrieval and QA performance on in-domain datasets, matching or even slightly outperforming state-of-the-art dense retrievers and readers. Moreover, end-to-end adaptation of our model significantly boosts its performance on out-of-domain datasets in both supervised and unsupervised settings, making our model a simple and adaptable end-to-end solution for knowledge-intensive tasks.

Outside-Knowledge Visual Question Answering (OK-VQA) is a challenging VQA task that requires retrieval of external knowledge to answer questions about images. Recent OK-VQA systems use Dense Passage Retrieval (DPR) to retrieve documents from external knowledge bases, such as Wikipedia, but with DPR trained separately from answer generation, introducing a potential limit on the overall system performance.Instead, we propose a joint training scheme which includes differentiable DPR integrated with answer generation so that the system can be trained in an end-to-end fashion. Our experiments show that our scheme outperforms recent OK-VQA systems with strong DPR for retrieval. We also introduce new diagnostic metrics to analyze how retrieval and generation interact. The strong retrieval ability of our model significantly reduces the number of retrieved documents needed in training, yielding significant benefits in answer quality and computation required for training.

A common thread of retrieval-augmented methods in the existing literature focuses on retrieving encyclopedic knowledge, such as Wikipedia, which facilitates well-defined entity and relation spaces that can be modeled. However, applying such methods to commonsense reasoning tasks faces two unique challenges, i.e., the lack of a general large-scale corpus for retrieval and a corresponding effective commonsense retriever. In this paper, we systematically investigate how to leverage commonsense knowledge retrieval to improve commonsense reasoning tasks. We proposed a unified framework of retrieval-augmented commonsense reasoning (called RACo), including a newly constructed commonsense corpus with over 20 million documents and novel strategies for training a commonsense retriever. We conducted experiments on four different commonsense reasoning tasks. Extensive evaluation results showed that our proposed RACo can significantly outperform other knowledge-enhanced method counterparts, achieving new SoTA performance on the CommonGen and CREAK leaderboards.

Temporal sentence grounding (TSG) aims to identify the temporal boundary of a specific segment from an untrimmed video by a sentence query. All existing works first utilize a sparse sampling strategy to extract a fixed number of video frames and then interact them with query for reasoning.However, we argue that these methods have overlooked two indispensable issues:1) Boundary-bias: The annotated target segment generally refers to two specific frames as corresponding start and end timestamps. The video downsampling process may lose these two frames and take the adjacent irrelevant frames as new boundaries.2) Reasoning-bias: Such incorrect new boundary frames also lead to the reasoning bias during frame-query interaction, reducing the generalization ability of model.To alleviate above limitations, in this paper, we propose a novel Siamese Sampling and Reasoning Network (SSRN) for TSG, which introduces a siamese sampling mechanism to generate additional contextual frames to enrich and refine the new boundaries. Specifically, a reasoning strategy is developed to learn the inter-relationship among these frames and generate soft labels on boundaries for more accurate frame-query reasoning. Such mechanism is also able to supplement the absent consecutive visual semantics to the sampled sparse frames for fine-grained activity understanding.Extensive experiments demonstrate the effectiveness of SSRN on three challenging datasets.

Large language models (LMs) are able to in-context learn—perform a new task via inference alone by conditioning on a few input-label pairs (demonstrations) and making predictions for new inputs. However, there has been little understanding of how the model learns and which aspects of the demonstrations contribute to end task performance. In this paper, we show that ground truth demonstrations are in fact not required—randomly replacing labels in the demonstrations barely hurts performance on a range of classification and multi-choce tasks, consistently over 12 different models including GPT-3. Instead, we find that other aspects of the demonstrations are the key drivers of endtask performance, including the fact that they provide a few examples of (1) the label space, (2) the distribution of the input text, and (3) the overall format of the sequence. Together, our analysis provides a new way of understanding how and why in-context learning works, while opening up new questions about how much can be learned from large language models through inference alone.

One of the main drivers of the recent advances in authorship verification is the PAN large-scale authorship dataset. Despite generating significant progress in the field, inconsistent performance differences between the closed and open test sets have been reported. To this end, we improve the experimental setup by proposing five new public splits over the PAN dataset, specifically designed to isolate and identify biases related to the text topic and to the author’s writing style. We evaluate several BERT-like baselines on these splits, showing that such models are competitive with authorship verification state-of-the-art methods. Furthermore, using explainable AI, we find that these baselines are biased towards named entities. We show that models trained without the named entities obtain better results and generalize better when tested on DarkReddit, our new dataset for authorship verification.

Knowledge distillation has been proven effective when customizing small language models for specific tasks. Here, a corpus as ‘textbook’ plays an indispensable role, only through which the teacher can teach the student. Prevailing methods adopt a two-stage distillation paradigm: general distillation first with task-agnostic general corpus and task-specific distillation next with augmented task-specific corpus. We argue that such a paradigm may not be optimal. In general distillation, it’s extravagant to let the diverse but desultory general knowledge overwhelms the limited model capacity of the student. While in task-specific distillation, the task corpus is usually limited and narrow, preventing the student from learning enough knowledge. To mitigate the issues in the two gapped corpora, we present a better textbook for the student to learn: contextualized corpus that contextualizes task corpus with large-scale general corpus through relevance-based text retrieval. Experimental results on GLUE benchmark demonstrate that contextualized corpus is the better textbook compared with jointly using general corpus and augmented task-specific corpus. Surprisingly, it enables task-specific distillation from scratch without general distillation while maintaining comparable performance, making it more flexible to customize the student model with desired model size under various computation constraints.