A Survey of Knowledge-Intensive NLP with Pre-Trained Language Models

With the increasing of model capacity brought by pre-trained language models, there emerges boosting needs for more knowledgeable natural language processing (NLP) models with advanced functionalities including providing and making flexible use of encyclopedic and commonsense knowledge. The mere pre-trained language models, however, lack the capacity of handling such knowledge-intensive NLP tasks alone. To address this challenge, large numbers of pre-trained language models augmented with external knowledge sources are proposed and in rapid development. In this paper, we aim to summarize the current progress of pre-trained language model-based knowledge-enhanced models (PLMKEs) by dissecting their three vital elements: knowledge sources, knowledge-intensive NLP tasks, and knowledge fusion methods. Finally, we present the challenges of PLMKEs based on the discussion regarding the three elements and attempt to provide NLP practitioners with potential directions for further research.

Knowledge-Rich Self-Supervised Entity Linking

Entity linking faces significant challenges, such as prolific variations and prevalent ambiguities, especially in high-value domains with myriad entities. Standard classification approaches suffer from the annotation bottleneck and cannot effectively handle unseen entities. Zero-shot entity linking has emerged as a promising direction for generalizing to new entities, but it still requires example gold entity mentions during training and canonical descriptions for all entities, both of which are rarely available outside of Wikipedia. In this paper, we explore Knowledge-RIch Self-Supervision ($\tt KRISS$) for entity linking, by leveraging readily available domain knowledge. In training, it generates self-supervised mention examples on unlabeled text using a domain ontology and trains a contextual encoder using contrastive learning. For inference, it samples self-supervised mentions as prototypes for each entity and conducts linking by mapping the test mention to the most similar prototype. Our approach subsumes zero-shot and few-shot methods, and can easily incorporate entity descriptions and gold mention labels if available. Using biomedicine as a case study, we conducted extensive experiments on seven standard datasets spanning biomedical literature and clinical notes. Without using any labeled information, our method produces $\tt KRISSBERT$, a universal entity linker for four million UMLS entities, which attains new state of the art, outperforming prior self-supervised methods by as much as over 20 absolute points in accuracy.

Fine-Tuning Large Neural Language Models for Biomedical Natural Language Processing

Motivation: A perennial challenge for biomedical researchers and clinical practitioners is to stay abreast with the rapid growth of publications and medical notes. Natural language processing (NLP) has emerged as a promising direction for taming information overload. In particular, large neural language models facilitate transfer learning by pretraining on unlabeled text, as exemplified by the successes of BERT models in various NLP applications. However, fine-tuning such models for an end task remains challenging, especially with small labeled datasets, which are common in biomedical NLP. Results: We conduct a systematic study on fine-tuning stability in biomedical NLP. We show that finetuning performance may be sensitive to pretraining settings, especially in low-resource domains. Large models have potential to attain better performance, but increasing model size also exacerbates finetuning instability. We thus conduct a comprehensive exploration of techniques for addressing fine-tuning instability. We show that these techniques can substantially improve fine-tuning performance for lowresource biomedical NLP applications. Specifically, freezing lower layers is helpful for standard BERT-BASE models, while layerwise decay is more effective for BERT-LARGE and ELECTRA models. For low-resource text similarity tasks such as BIOSSES, reinitializing the top layer is the optimal strategy. Overall, domainspecific vocabulary and pretraining facilitate more robust models for fine-tuning. Based on these findings, we establish new state of the art on a wide range of biomedical NLP applications. Availability and implementation: To facilitate progress in biomedical NLP, we release our state-of-the-art pretrained and fine-tuned models: https://aka.ms/BLURB.

Human Parity on CommonsenseQA: Augmenting Self-Attention with External Attention

Most of today's AI systems focus on using self-attention mechanisms and transformer architectures on large amounts of diverse data to achieve impressive performance gains. In this paper, we propose to augment the transformer architecture with an external attention mechanism to bring external knowledge and context to bear. By integrating external information into the prediction process, we hope to reduce the need for ever-larger models and increase the democratization of AI systems. We find that the proposed external attention mechanism can significantly improve the performance of existing AI systems, allowing practitioners to easily customize foundation AI models to many diverse downstream applications. In particular, we focus on the task of Commonsense Reasoning, demonstrating that the proposed external attention mechanism can augment existing transformer models and significantly improve the model's reasoning capabilities. The proposed system, Knowledgeable External Attention for commonsense Reasoning (KEAR), reaches human parity on the open CommonsenseQA research benchmark with an accuracy of 89.4\% in comparison to the human accuracy of 88.9\%.

Two Wrongs Don't Make a Right: Combating Confirmation Bias in Learning with Label Noise

Noisy labels damage the performance of deep networks. For robust learning, a prominent two-stage pipeline alternates between eliminating possible incorrect labels and semi-supervised training. However, discarding part of observed labels could result in a loss of information, especially when the corruption is not completely random, e.g., class-dependent or instance-dependent. Moreover, from the training dynamics of a representative two-stage method DivideMix, we identify the domination of confirmation bias: Pseudo-labels fail to correct a considerable amount of noisy labels and consequently, the errors accumulate. To sufficiently exploit information from observed labels and mitigate wrong corrections, we propose Robust Label Refurbishment (Robust LR)-a new hybrid method that integrates pseudo-labeling and confidence estimation techniques to refurbish noisy labels. We show that our method successfully alleviates the damage of both label noise and confirmation bias. As a result, it achieves state-of-the-art results across datasets and noise types. For example, Robust LR achieves up to 4.5% absolute top-1 accuracy improvement over the previous best on the real-world noisy dataset WebVision.

CLUES: Few-Shot Learning Evaluation in Natural Language Understanding

Most recent progress in natural language understanding (NLU) has been driven, in part, by benchmarks such as GLUE, SuperGLUE, SQuAD, etc. In fact, many NLU models have now matched or exceeded "human-level" performance on many tasks in these benchmarks. Most of these benchmarks, however, give models access to relatively large amounts of labeled data for training. As such, the models are provided far more data than required by humans to achieve strong performance. That has motivated a line of work that focuses on improving few-shot learning performance of NLU models. However, there is a lack of standardized evaluation benchmarks for few-shot NLU resulting in different experimental settings in different papers. To help accelerate this line of work, we introduce CLUES (Constrained Language Understanding Evaluation Standard), a benchmark for evaluating the few-shot learning capabilities of NLU models. We demonstrate that while recent models reach human performance when they have access to large amounts of labeled data, there is a huge gap in performance in the few-shot setting for most tasks. We also demonstrate differences between alternative model families and adaptation techniques in the few shot setting. Finally, we discuss several principles and choices in designing the experimental settings for evaluating the true few-shot learning performance and suggest a unified standardized approach to few-shot learning evaluation. We aim to encourage research on NLU models that can generalize to new tasks with a small number of examples. Code and data for CLUES are available at https://github.com/microsoft/CLUES.

RMNet: Equivalently Removing Residual Connection from Networks

Although residual connection enables training very deep neural networks, it is not friendly for online inference due to its multi-branch topology. This encourages many researchers to work on designing DNNs without residual connections at inference. For example, RepVGG re-parameterizes multi-branch topology to a VGG-like (single-branch) model when deploying, showing great performance when the network is relatively shallow. However, RepVGG can not transform ResNet to VGG equivalently because re-parameterizing methods can only be applied to linear blocks and the non-linear layers (ReLU) have to be put outside of the residual connection which results in limited representation ability, especially for deeper networks. In this paper, we aim to remedy this problem and propose to remove the residual connection in a vanilla ResNet equivalently by a reserving and merging (RM) operation on ResBlock. Specifically, the RM operation allows input feature maps to pass through the block while reserving their information and merges all the information at the end of each block, which can remove residual connections without changing the original output. As a plug-in method, RM Operation basically has three advantages: 1) its implementation makes it naturally friendly for high ratio network pruning. 2) it helps break the depth limitation of RepVGG. 3) it leads to better accuracy-speed trade-off network (RMNet) compared to ResNet and RepVGG. We believe the ideology of RM Operation can inspire many insights on model design for the community in the future. Code is available at: https://github.com/fxmeng/RMNet.

Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotations

Existing research on learning with noisy labels mainly focuses on synthetic label noise. Synthetic label noise, though has clean structures which greatly enable statistical analyses, often fails to model the real-world noise patterns. The recent literature has observed several efforts to offer real-world noisy datasets, yet the existing efforts suffer from two caveats: firstly, the lack of ground-truth verification makes it hard to theoretically study the property and treatment of real-world label noise. Secondly, these efforts are often of large scales, which may lead to unfair comparisons of robust methods within reasonable and accessible computation power. To better understand real-world label noise, it is important to establish controllable and moderate-sized real-world noisy datasets with both ground-truth and noisy labels. This work presents two new benchmark datasets (CIFAR-10N, CIFAR-100N), equipping the train dataset of CIFAR-10 and CIFAR-100 with human-annotated real-world noisy labels that we collect from Amazon Mechanical Turk. We quantitatively and qualitatively show that real-world noisy labels follow an instance-dependent pattern rather than the classically adopted class-dependent ones. We then initiate an effort to benchmark a subset of existing solutions using CIFAR-10N, CIFAR-100N. We next proceed to study the memorization of model predictions, which further illustrates the difference between human noise and class-dependent synthetic noise. We show indeed the real-world noise patterns impose new and outstanding challenges as compared to synthetic ones. These observations require us to rethink the treatment of noisy labels, and we hope the availability of these two datasets would facilitate the development and evaluation of future learning with noisy label solutions. The corresponding datasets and the leaderboard are publicly available at \url{http://noisylabels.com}.

Demystifying How Self-Supervised Features Improve Training from Noisy Labels

The advancement of self-supervised learning (SSL) motivates researchers to apply SSL on other tasks such as learning with noisy labels. Recent literature indicates that methods built on SSL features can substantially improve the performance of learning with noisy labels. Nonetheless, the deeper reasons why (and how) SSL features benefit the training from noisy labels are less understood. In this paper, we study why and how self-supervised features help networks resist label noise using both theoretical analyses and numerical experiments. Our result shows that, given a quality encoder pre-trained from SSL, a simple linear layer trained by the cross-entropy loss is theoretically robust to symmetric label noise. Further, we provide insights for how knowledge distilled from SSL features can alleviate the over-fitting problem. We hope our work provides a better understanding for learning with noisy labels from the perspective of self-supervised learning and can potentially serve as a guideline for further research. Code is available at github.com/UCSC-REAL/SelfSup_NoisyLabel.