Causal transformer language models (LMs), such as GPT-3, typically require some form of positional encoding, such as positional embeddings. However, we show that LMs without any explicit positional encoding are still competitive with standard models, and that this phenomenon is robust across different datasets, model sizes, and sequence lengths. Probing experiments reveal that such models acquire an implicit notion of absolute positions throughout the network, effectively compensating for the missing information. We conjecture that causal attention enables the model to infer the number of predecessors that each token can attend to, thereby approximating its absolute position. Our findings indicate that causal LMs might derive positional awareness not only from the explicit positioning mechanism, but also from the effects of the causal mask.

We introduce Transformer Grammars (TGs), a novel class of Transformer language models that combine (i) the expressive power, scalability, and strong performance of Transformers and (ii) recursive syntactic compositions, which here are implemented through a special attention mask and deterministic transformation of the linearized tree. We find that TGs outperform various strong baselines on sentence-level language modeling perplexity, as well as on multiple syntax-sensitive language modeling evaluation metrics. Additionally, we find that the recursive syntactic composition bottleneck which represents each sentence as a single vector harms perplexity on document-level language modeling, providing evidence that a different kind of memory mechanism -- one that is independent of composed syntactic representations -- plays an important role in current successful models of long text.

The attention mechanism is considered the backbone of the widely-used Transformer architecture. It contextualizes the input by computing input-specific attention matrices. We find that this mechanism, while powerful and elegant, is not as important as typically thought for pretrained language models. We introduce PAPA, a new probing method that replaces the input-dependent attention matrices with constant ones -- the average attention weights over multiple inputs. We use PAPA to analyze several established pretrained Transformers on six downstream tasks. We find that without any input-dependent attention, all models achieve competitive performance -- an average relative drop of only 8% from the probing baseline. Further, little or no performance drop is observed when replacing half of the input-dependent attention matrices with constant (input-independent) ones. Interestingly, we show that better-performing models lose more from applying our method than weaker models, suggesting that the utilization of the input-dependent attention mechanism might be a factor in their success. Our results motivate research on simpler alternatives to input-dependent attention, as well as on methods for better utilization of this mechanism in the Transformer architecture.

标准预审进的语言模型可在子字代币序列上运行，而无需直接访问组成每个令牌字符串表示的字符。我们探究了预审前的语言模型的嵌入层，并表明模型在一个令人惊讶的程度上学习了整个单词和子字代币的内部字符组成，而没有看到字符和令牌。我们的结果表明，罗伯塔（Roberta）的嵌入层具有足够的信息，可以准确地阐明词汇的三分之一，并在所有令牌类型上达到高平均角色Ngram重叠。我们进一步测试了使用其他字符信息丰富子词模型是否可以改善语言建模，并观察到该方法具有几乎相同的学习曲线，作为训练而无需基于拼写的丰富。总体而言，我们的结果表明，语言建模目标激励模型隐式学习一些拼写概念，并且明确教授模型如何拼写的方式似乎并没有增强其在此类任务上的绩效。

Recent progress in pre-trained neural language models has significantly improved the performance of many natural language processing (NLP) tasks. In this paper we propose a new model architecture DeBERTa (Decoding-enhanced BERT with disentangled attention) that improves the BERT and RoBERTa models using two novel techniques. The first is the disentangled attention mechanism, where each word is represented using two vectors that encode its content and position, respectively, and the attention weights among words are computed using disentangled matrices on their contents and relative positions, respectively. Second, an enhanced mask decoder is used to incorporate absolute positions in the decoding layer to predict the masked tokens in model pre-training. In addition, a new virtual adversarial training method is used for fine-tuning to improve models' generalization. We show that these techniques significantly improve the efficiency of model pre-training and the performance of both natural language understand (NLU) and natural langauge generation (NLG) downstream tasks. Compared to RoBERTa-Large, a DeBERTa model trained on half of the training data performs consistently better on a wide range of NLP tasks, achieving improvements on MNLI by +0.9% (90.2% vs. 91.1%), on SQuAD v2.0 by +2.3% (88.4% vs. 90.7%) and RACE by +3.6% (83.2% vs. 86.8%). Notably, we scale up DeBERTa by training a larger version that consists of 48 Transform layers with 1.5 billion parameters. The significant performance boost makes the single DeBERTa model surpass the human performance on the SuperGLUE benchmark (Wang et al., 2019a) for the first time in terms of macro-average score (89.9 versus 89.8), and the ensemble DeBERTa model sits atop the SuperGLUE leaderboard as of January 6, 2021, outperforming the human baseline by a decent margin (90.3 versus 89.8). The pre-trained DeBERTa models and the source code were released at: https://github.com/microsoft/DeBERTa 1 .

We introduce a new language representation model called BERT, which stands for Bidirectional Encoder Representations from Transformers. Unlike recent language representation models (Peters et al., 2018a;Radford et al., 2018), BERT is designed to pretrain deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers. As a result, the pre-trained BERT model can be finetuned with just one additional output layer to create state-of-the-art models for a wide range of tasks, such as question answering and language inference, without substantial taskspecific architecture modifications.BERT is conceptually simple and empirically powerful. It obtains new state-of-the-art results on eleven natural language processing tasks, including pushing the GLUE score to 80.5% (7.7% point absolute improvement), MultiNLI accuracy to 86.7% (4.6% absolute improvement), SQuAD v1.1 question answering Test F1 to 93.2 (1.5 point absolute improvement) and SQuAD v2.0 Test F1 to 83.1 (5.1 point absolute improvement).

许多NLP任务需要处理超出预磨模模型的长度限制的长语境。为了将这些模型扩展到更长的文本序列，已经提出了许多有效的远程注意力变体。尽管沿着这个方向进行了丰富的研究，但仍然难以在实际用例中衡量这些模型的相对有效性，例如，如果我们在预先rain-yfetune范式之后应用这些模型。在这项工作中，我们的目标是对这些具有大规模和受控实验的这些新兴模型进行彻底的分析。对于每个关注变体，我们使用相同的长DOC语料库，然后使用相同的长DOC语料库，然后为现实世界的长情节任务进行芬特这些模型。我们的调查结果揭示了现有广泛使用的远程基准的陷阱，并显示任何经过测试的高效关注可以在标准预介质范式下击败一个简单的本地窗口关注。对本地注意力变化的进一步分析表明，即使是常用的注意力窗口重叠也没有必要实现良好的下游结果 - 使用不相交的本地关注，我们能够构建符合性能的更简单且更高效的Long-Doc QA模型霍尔福勒〜\ citep {longformer}其预先花费的一半。

变压器注意机制的二次计算和内存复杂性限制了对长序列建模的可扩展性。在本文中，我们提出了Luna，一种线性统一嵌套关注机制，使Softmax注意力具有两个嵌套线性关注功能，仅产生线性（与二次）的时间和空间复杂度相反。具体地，通过第一注意功能，LUNA将输入序列包装成固定长度的序列。然后，使用第二关注功能未包装包装序列。与更传统的关注机制相比，LUNA引入具有固定长度的附加序列作为输入和额外的相应输出，允许LUNA线性地进行关注操作，同时还存储足够的上下文信息。我们对三个序列建模任务的基准进行了广泛的评估：长上下文序列建模，神经机平移和大型预磨损的屏蔽语言建模。竞争甚至更好的实验结果表明了Luna的有效性和效率与各种各样相比

在预介质期间，预解压器变压器遭受梯度幅度不匹配：早期层处的梯度远远大于更高层的层。我们所提出的常规程序架构可以减轻这些问题，这为每层增加了三个归一化操作：自我注意后的一层规范，自我注意输出的头部明智的缩放，以及第一完全连接层之后的层标。额外的运营产生忽略不计的计算成本（+ 0.4％的参数增加），但是改善了从12500万到27亿个参数的因果和屏蔽语言模型的预先欣赏困惑和下游任务性能。例如，在我们最强的1.3B参数基线顶部添加NARMFORMER可以在相同的计算预算中更快地达到24％的平等困惑，或者更好地收敛0.27困惑。该模型达到GPT3大（1.3B）零拍摄性能速度快60％。对于屏蔽语言建模，Normformer平均将微调胶水性能提高1.9％。 Fairseq HTTPS://github.com/pytorch/faireq/tree/main/examples/normformer提供培训ormalformer模型的代码。

变压器模型是置换等分之一的。要提供输入令牌的顺序和类型信息，通常将位置和段嵌入式添加到输入中。最近的作品提出了具有相对位置编码的位置编码的变化，实现了更好的性能。我们的分析表明，增益实际上来自从输入中将位置信息移动到注意层。由此激励，我们介绍了变压器（饮食）的解耦的位置注意，一个简单但有效的机制，将位置和分段信息编码为变压器模型。该方法具有更快的培训和推理时间，同时在胶水，Xtreme和WMT基准上实现竞争性能。我们进一步概括了我们的方法到远程变压器并显示性能增益。

Position modeling plays a critical role in Transformers. In this paper, we focus on length extrapolation, i.e., training on short texts while evaluating longer sequences. We define attention resolution as an indicator of extrapolation. Then we propose two designs to improve the above metric of Transformers. Specifically, we introduce a relative position embedding to explicitly maximize attention resolution. Moreover, we use blockwise causal attention during inference for better resolution. We evaluate different Transformer variants with language modeling. Experimental results show that our model achieves strong performance in both interpolation and extrapolation settings. The code will be available at https://aka.ms/LeX-Transformer.

The pre-training of masked language models (MLMs) consumes massive computation to achieve good results on downstream NLP tasks, resulting in a large carbon footprint. In the vanilla MLM, the virtual tokens, [MASK]s, act as placeholders and gather the contextualized information from unmasked tokens to restore the corrupted information. It raises the question of whether we can append [MASK]s at a later layer, to reduce the sequence length for earlier layers and make the pre-training more efficient. We show: (1) [MASK]s can indeed be appended at a later layer, being disentangled from the word embedding; (2) The gathering of contextualized information from unmasked tokens can be conducted with a few layers. By further increasing the masking rate from 15% to 50%, we can pre-train RoBERTa-base and RoBERTa-large from scratch with only 78% and 68% of the original computational budget without any degradation on the GLUE benchmark. When pre-training with the original budget, our method outperforms RoBERTa for 6 out of 8 GLUE tasks, on average by 0.4%.

我们通过与与前面令牌的局部相似度，通过调节从大语料库检索的文档块来增强自动回归语言模型。尽管使用25美元\时分，我们的检索增强型变压器（RetroCro）的检索增强型变压器（RetroCr）对GPT-3和侏罗纪-1获得了可比性的性能。微调后，复古表演转换为下游知识密集型任务，如问题应答。复古结合了冷冻BERT猎犬，一种可微分的编码器和块状的横向机制，以预测基于数量级的令牌，而不是训练期间通常消耗的数量。我们通常从头开始训练复古，还可以快速改造预先接受的变压器，通过检索，仍然达到良好的性能。我们的工作通过以前所未有的规模开辟了通过显式内存改进语言模型的新途径。

Transformer networks have seen great success in natural language processing and machine vision, where task objectives such as next word prediction and image classification benefit from nuanced context sensitivity across high-dimensional inputs. However, there is an ongoing debate about how and when transformers can acquire highly structured behavior and achieve systematic generalization. Here, we explore how well a causal transformer can perform a set of algorithmic tasks, including copying, sorting, and hierarchical compositions of these operations. We demonstrate strong generalization to sequences longer than those used in training by replacing the standard positional encoding typically used in transformers with labels arbitrarily paired with items in the sequence. We search for the layer and head configuration sufficient to solve these tasks, then probe for signs of systematic processing in latent representations and attention patterns. We show that two-layer transformers learn reliable solutions to multi-level problems, develop signs of task decomposition, and encode input items in a way that encourages the exploitation of shared computation across related tasks. These results provide key insights into how attention layers support structured computation both within a task and across multiple tasks.

Transformer language models (TLMs) are critical for most NLP tasks, but they are difficult to create for low-resource languages because of how much pretraining data they require. In this work, we investigate two techniques for training monolingual TLMs in a low-resource setting: greatly reducing TLM size, and complementing the masked language modeling objective with two linguistically rich supervised tasks (part-of-speech tagging and dependency parsing). Results from 7 diverse languages indicate that our model, MicroBERT, is able to produce marked improvements in downstream task evaluations relative to a typical monolingual TLM pretraining approach. Specifically, we find that monolingual MicroBERT models achieve gains of up to 18% for parser LAS and 11% for NER F1 compared to a multilingual baseline, mBERT, while having less than 1% of its parameter count. We conclude reducing TLM parameter count and using labeled data for pretraining low-resource TLMs can yield large quality benefits and in some cases produce models that outperform multilingual approaches.

基于变压器的模型在多个领域和任务上显示了它们的有效性。自我注意力允许将所有序列元素的信息结合到上下文感知表示形式中。但是，全球和本地信息必须主要存储在相同的元素表示中。此外，输入序列的长度受到自我注意的二次计算复杂性的限制。在这项工作中，我们提出并研究了一个记忆启动的片段级循环变压器（复发记忆变压器）。内存允许借助复发的帮助存储和处理本地和全局信息，并可以在长序列的段之间传递信息。我们通过将特殊的内存令牌添加到输入或输出序列中，实现了一个内存机制，无需更改变压器模型。然后，对变压器进行了训练，以控制内存操作和序列表示处理。实验的结果表明，我们的模型与Transformer-XL在语言建模上的较小内存大小上的表现相同，并在需要更长序列处理的任务方面胜过它。我们证明，将内存令牌添加到TR-XL可以提高IT性能。这使得反复的内存变压器成为需要学习长期依赖性和内存处理中的通用性（例如算法任务和推理）的应用程序的有前途的体系结构。

With the capability of modeling bidirectional contexts, denoising autoencoding based pretraining like BERT achieves better performance than pretraining approaches based on autoregressive language modeling. However, relying on corrupting the input with masks, BERT neglects dependency between the masked positions and suffers from a pretrain-finetune discrepancy. In light of these pros and cons, we propose XLNet, a generalized autoregressive pretraining method that (1) enables learning bidirectional contexts by maximizing the expected likelihood over all permutations of the factorization order and (2) overcomes the limitations of BERT thanks to its autoregressive formulation. Furthermore, XLNet integrates ideas from Transformer-XL, the state-of-the-art autoregressive model, into pretraining. Empirically, under comparable experiment settings, XLNet outperforms BERT on 20 tasks, often by a large margin, including question answering, natural language inference, sentiment analysis, and document ranking. 1 .

最近，大型预用语言模型（LMS）越来越受欢迎。培训这些模型需要更多的计算资源，并且大多数现有模型仅在英文文本上培训。以其他语言训练这些模型非常昂贵。为了缓解这个问题，我们介绍了一种叫做威施塞的方法 - 将英语模型传输到新语言。我们将英语模型的销量与目标语言中的销量交换，并初始化令牌嵌入式，以便通过利用覆盖英语和目标语言的多语言静态字嵌入来初始化令牌嵌入式。我们使用Wechsel将GPT-2和Roberta模型转移到4种其他语言（法语，德语，中文和斯瓦希里语）。 Wechsel通过以前提出的跨语言参数转移和优于比较大小的模型来改善从目标语言的划痕训练的相当大小的型号，距离培训速度较小。我们的方法使培训大型语言模型为新语言更容易访问，更少损害环境。我们宣传我们的代码和型号。

尽管蒙版语言模型具有高度性能，并且被NLP从业人员广泛采用，但它们不能轻易用于自回归语言建模（下一个单词预测和序列概率估计）。我们提出了一种基于LSTM的自回归语言模型，该模型使用融合（例如串联）使用前缀嵌入（来自验证的蒙版语言模型），以获得语言建模的更丰富的上下文表示。我们发现Fusion有助于可靠地降低困惑（16.74 $ \ rightarrow $ 15.80），甚至在从与培训数据的不同领域传输到数据集后，它甚至可以保留。我们还通过将其下一个单词的惊人估计与人类阅读时间相关联，评估了表现最佳的融合模型。与我们的期望相矛盾，尽管整体上的困惑程度有所改善，但相关性仍然与基线模型相同。最后，尽管我们专注于在文本上预先训练的语言模型作为融合的来源，但我们的方法可能会扩展到将表示为固定尺寸矢量表示的任何信息融合到自动回归语言模型中。这些包括例如句子外部信息是为知识库或多模式编码器的表示形式检索的。

我们在变压器中重新审视设计选择，并提出方法来解决它们在处理长序列中的弱点。首先，我们提出了一个名为“门控注意单元”的简单层，该层允许使用较弱的单头注意，而质量损失最小。然后，我们提出了一种与该新层的线性近似方法互补的，该方法对加速器友好且质量高度竞争。最终的型号（名为Flash）与短（512）和长（8K）上下文长度相匹配，在WIKI-40B上达到高达4.9 $ \ times $的训练速度和PG上的12.1 $ \ times $，在PG上达到了4.9 $ \ times $的困惑。-19用于自动回归语言建模，C4的4.8 $ \ times $用于掩盖语言建模。