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.
Center for Information and Language Processing（信息与语言处理中心）；Ludwig Maximilian University of Munich（慕尼黑大学）
This paper describes how to apply self-attention with relative positional encodings to the task of relation extraction. We propose to use the self-attention encoder layer together with an additional position-aware attention layer that takes into account positions of the query and the object in the sentence. The self-attention encoder also uses a custom implementation of relative positional encodings which allow each word in the sentence to take into account its left and right context. The evaluation of the model is done on the TACRED dataset. The proposed model relies only on attention (no rec…
IIIT Sri City（印度信息技术学院斯里城）；UCLA（加州大学洛杉矶分校）；Wipro AI Labs（威普罗人工智能实验室）；AI Institute, University of South Carolina（南卡罗来纳大学人工智能研究所）
NLP applications for code-mixed (CM) or mix-lingual text have gained a significant momentum recently, the main reason being the prevalence of language mixing in social media communications in multi-lingual societies like India, Mexico, Europe, parts of USA etc. Word embeddings are basic build-ing blocks of any NLP system today, yet, word embedding for CM languages is an unexplored territory. The major bottleneck for CM word embeddings is switching points, where the language switches. These locations lack in contextually and statistical systems fail to model this phenomena due to high variance …
Transformer models are permutation equivariant. To supply the order and type information of the input tokens, position and segment embeddings are usually added to the input. Recent works proposed variations of positional encodings with relative position encodings achieving better performance. Our analysis shows that the gain actually comes from moving positional information to attention layer from the input. Motivated by this, we introduce Decoupled Positional Attention for Transformers (DIET), a simple yet effective mechanism to encode position and segment information into the Transformer mod…
University of Oslo（奥斯陆大学）；University of Copenhagen（哥本哈根大学）
In order to preserve word-order information in a non-autoregressive setting, transformer architectures tend to include positional knowledge, by (for instance) adding positional encodings to token embeddings. Several modifications have been proposed over the sinusoidal positional encodings used in the original transformer architecture; these include, for instance, separating position encodings and token embeddings, or directly modifying attention weights based on the distance between word pairs. We first show that surprisingly, while these modifications tend to improve monolingual language mode…
Bayesian Attention Mechanism: A Probabilistic Framework for Positional Encoding and Context Length Extrapolation
MALTA – Machine Learning Theory and Applications Lab, PUCRS – Porto Alegre, Brazil（MALTA机器学习理论与应用实验室，PUCRS-波士顿实验室，巴西）；Kunumi Institute, Brazil（昆米研究所，巴西）
One key challenge in multi-document summarization is to capture the relations among input documents that distinguish between single document summarization (SDS) and multi-document summarization (MDS). Few existing MDS works address this issue. One effective way is to encode document positional information to assist models in capturing cross-document relations. However, existing MDS models, such as Transformer-based models, only consider token-level positional information. Moreover, these models fail to capture sentences' linguistic structure, which inevitably causes confusions in the generated…
In this work, we investigate the positional encoding methods used in language pre-training (e.g., BERT) and identify several problems in the existing formulations. First, we show that in the absolute positional encoding, the addition operation applied on positional embeddings and word embeddings brings mixed correlations between the two heterogeneous information resources. It may bring unnecessary randomness in the attention and further limit the expressiveness of the model. Second, we question whether treating the position of the symbol \texttt{[CLS]} the same as other words is a reasonable d…
Google Cloud AI（谷歌云人工智能）；McGill University（麦吉尔大学）；Google Research（谷歌研究院）
Natural reading orders of words are crucial for information extraction from form-like documents. Despite recent advances in Graph Convolutional Networks (GCNs) on modeling spatial layout patterns of documents, they have limited ability to capture reading orders of given word-level node representations in a graph. We propose Reading Order Equivariant Positional Encoding (ROPE), a new positional encoding technique designed to apprehend the sequential presentation of words in documents. ROPE generates unique reading order codes for neighboring words relative to the target word given a word-level …
CONFLATOR: Incorporating Switching Point based Rotatory Positional Encodings for Code-Mixed Language Modeling
San Diego State University（圣地亚哥州立大学）；San Jose State University（圣何塞州立大学）；Wipro AI Labs（威普罗人工智能实验室）；University of California Los Angeles（加利福尼亚大学洛杉矶分校）；Stanford University（斯坦福大学）；University of South Carolina（南卡罗来纳大学）；Amazon AI（亚马逊人工智能）
Shanghai AI Laboratory（上海人工智能实验室）；Australian National University（澳大利亚国立大学）；Northwestern Polytechnical University（西北工业大学）；The University of Hong Kong（香港大学）
Auto-regressive neural sequence models have been shown to be effective across text generation tasks. However, their left-to-right decoding order prevents generation from being parallelized. Insertion Transformer (Stern et al., 2019) is an attractive alternative that allows outputting multiple tokens in a single generation step. Nevertheless, due to the incompatibility between absolute positional encoding and insertion-based generation schemes, it needs to refresh the encoding of every token in the generated partial hypothesis at each step, which could be costly. We design a novel reusable posi…
Apple（苹果公司）；Facebook AI Research（Facebook 人工智能研究院）；Herophilus, Inc.（Herophilus 公司）
Without positional information, attention-based Transformer neural networks are permutation-invariant. Absolute or relative positional embeddings are the most popular ways to feed Transformer models with positional information. Absolute positional embeddings are simple to implement, but suffer from generalization issues when evaluating on sequences longer than seen at training time. Relative positions are more robust to input length change, but are more complex to implement and yield inferior model throughput due to extra computational and memory costs. In this paper, we propose an augmentatio…
The attention module, which is a crucial component in Transformer, cannot scale efficiently to long sequences due to its quadratic complexity. Many works focus on approximating the dot-then-exponentiate softmax function in the original attention, leading to sub-quadratic or even linear-complexity Transformer architectures. However, we show that these methods cannot be applied to more powerful attention modules that go beyond the dot-then-exponentiate style, e.g., Transformers with relative positional encoding (RPE). Since in many state-of-the-art models, relative positional encoding is used as…
In recent years, pre-trained Transformers have dominated the majority of NLP benchmark tasks. Many variants of pre-trained Transformers have kept breaking out, and most focus on designing different pre-training objectives or variants of self-attention. Embedding the position information in the self-attention mechanism is also an indispensable factor in Transformers however is often discussed at will. Therefore, this paper carries out an empirical study on position embeddings of mainstream pre-trained Transformers, which mainly focuses on two questions: 1) Do position embeddings really learn th…
Video-guided machine translation as one of multimodal neural machine translation tasks targeting on generating high-quality text translation by tangibly engaging both video and text. In this work, we presented our video-guided machine translation system in approaching the Video-guided Machine Translation Challenge 2020. This system employs keyframe-based video feature extractions along with the video feature positional encoding. In the evaluation phase, our system scored 36.60 corpus-level BLEU-4 and achieved the 1st place on the Video-guided Machine Translation Challenge 2020.
Inria（法国国家信息与自动化研究所）；Télécom Paris（巴黎电信学院）；Institut Polytechnique de Paris（巴黎理工学院）；Research Center for IT Innovation, Academia Sinica（中央研究院资讯科技创新研究中心）；National Taiwan University（台湾大学）；Taiwan AI Labs（台湾人工智能实验室）；PSL Research University（巴黎文理研究大学）；École Normale Supérieure（巴黎高等师范学院）
Recent advances in Transformer models allow for unprecedented sequence lengths, due to linear space and time complexity. In the meantime, relative positional encoding (RPE) was proposed as beneficial for classical Transformers and consists in exploiting lags instead of absolute positions for inference. Still, RPE is not available for the recent linear-variants of the Transformer, because it requires the explicit computation of the attention matrix, which is precisely what is avoided by such methods. In this paper, we bridge this gap and present Stochastic Positional Encoding as a way to genera…
Neural encoder-decoder models have been successful in natural language generation tasks. However, real applications of abstractive summarization must consider additional constraint that a generated summary should not exceed a desired length. In this paper, we propose a simple but effective extension of a sinusoidal positional encoding (Vaswani et al., 2017) to enable neural encoder-decoder model to preserves the length constraint. Unlike in previous studies where that learn embeddings representing each length, the proposed method can generate a text of any length even if the target length is n…
NTT Communication Science Laboratories（NTT通信科学实验室）；Nara Institute of Science and Technology（奈良先端科学技术大学院大学）
Non-autoregressive neural machine translation (NAT) usually employs sequence-level knowledge distillation using autoregressive neural machine translation (AT) as its teacher model. However, a NAT model often outputs shorter sentences than an AT model. In this work, we propose sequence-level knowledge distillation (SKD) using perturbed length-aware positional encoding and apply it to a student model, the Levenshtein Transformer. Our method outperformed a standard Levenshtein Transformer by 2.5 points in bilingual evaluation understudy (BLEU) at maximum in a WMT14 German to English translation. …
TU Dresden（德累斯顿工业大学）；Fraunhofer IAIS（弗劳恩霍夫智能分析与信息系统研究所）
While a considerable amount of semantic parsing approaches have employed RNN architectures for code generation tasks, there have been only few attempts to investigate the applicability of Transformers for this task. Including hierarchical information of the underlying programming language syntax has proven to be effective for code generation. Since the positional encoding of the Transformer can only represent positions in a flat sequence, we have extended the encoding scheme to allow the attention mechanism to also attend over hierarchical positions in the input. Furthermore, we have realized …
Mitigating Coordinate Prediction Bias from Positional Encoding Failures
The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；The Hong Kong University of Science and Technology（香港科技大学）；University of California, Los Angeles（加州大学洛杉矶分校）；The University of Queensland（昆士兰大学）；Shanghai University of Finance and Economics（上海财经大学）
Transformer has shown promising results in many sequence to sequence transformation tasks recently. It utilizes a number of feed-forward self-attention layers to replace the recurrent neural networks (RNN) in attention-based encoder decoder (AED) architecture. Self-attention layer learns temporal dependence by incorporating sinusoidal positional embedding of tokens in a sequence for parallel computing. Quicker iteration speed in training than sequential operation of RNN can be obtained. Deeper layers of the transformer also make it perform better than RNN-based AED. However, this parallelizati…
Karlsruhe Institute of Technology（卡尔斯鲁厄理工学院）；Johns Hopkins University（约翰斯·霍普金斯大学）；Maastricht University（马斯特里赫特大学）；Carnegie Mellon University（卡内基梅隆大学）
Transformer models are powerful sequence-to-sequence architectures that are capable of directly mapping speech inputs to transcriptions or translations. However, the mechanism for modeling positions in this model was tailored for text modeling, and thus is less ideal for acoustic inputs. In this work, we adapt the relative position encoding scheme to the Speech Transformer, where the key addition is relative distance between input states in the self-attention network. As a result, the network can better adapt to the variable distributions present in speech data. Our experiments show that our r…
