Transformer-XL: Attentive Language Models beyond a Fixed-Length Context
Open accessThe DOI matches its Crossref record; 65 of the 67 references it cites were found in a registry or an index.
Abstract
Transformers have a potential of learning longer-term dependency, but are limited by a fixed-length context in the setting of language modeling.We propose a novel neural architecture Transformer-XL that enables learning dependency beyond a fixed length without disrupting temporal coherence.It consists of a segment-level recurrence mechanism and a novel positional encoding scheme.Our method not only enables capturing longer-term dependency, but also resolves the context fragmentation problem.As a result, Transformer-XL learns dependency that is 80% longer than RNNs and 450% longer than vanilla Transformers, achieves better performance on both short and long sequences, and is up to 1,800+ times faster than vanilla Transformers during evaluation.Notably, we improve the state-ofthe-art results of bpc/perplexity to 0.99 on en-wiki8, 1.08 on text8, 18.3 on WikiText-103, 21.8 on One Billion Word, and 54.5 on Penn Treebank (without finetuning).When trained only on WikiText-103, Transformer-XL manages to generate reasonably coherent, novel text articles with thousands of tokens.Our code, pretrained models, and hyperparameters are available in both Tensorflow and PyTorch 1 .
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“note that the Transformer XL model25 incorporates some form of recurrent processing within the transformer block, but it still takes the entire text as its input in parallel.”
From Incremental accumulation of linguistic context in artificial and biological neural networks (2025) · cites this paper as [25]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Directly extending the context length of transformers incurs a quadratic increase in computational time and memory cost due to the transformer architecture's self-attention mechanism, making the design of new long-context architectures a pressing research challenge (Dai et al., 2019; Kitaev et al., 2020; Beltagy et al., 2020).”
From MemGPT: Towards LLMs as Operating Systems (2023) · cites this paper as Dai et al., 2019
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“We believe that the different trends are caused by the truncation problem [59], [60].”
From What Makes Good In-Context Demonstrations for Code Intelligence Tasks with LLMs? (2023) · cites this paper as [59]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Fixed-window methods : including context windows of 4K, 8K, and 16K tokens with FIFO eviction [10, 2]”
From FadeMem: Biologically-Inspired Forgetting for Efficient Agent Memory (2026) · cites this paper as [10]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Thus, in practice, Transformer-XL and Compressive Transformer need huge memory size to achieve good performance.”
From Memformer: a memory-augmented transformer for sequence modeling (2020) · cites this paper as Dai et al., 2019
Experimental — accuracy not yet measured
12 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 7×
“On the other hand, the direct connections between long-distance word pairs baked in attention mechanisms might ease optimization and enable the learning of long-term dependency (Bahdanau et al., 2014; Vaswani et al., 2017).”
BackgroundMethodInfluentialHochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 2×
“Recurrent neural networks (RNNs), in particular Long Short-Term Memory (LSTM) networks (Hochreiter & Schmidhuber, 1997), have been a standard solution to language modeling and obtained strong results on multiple benchmarks.”
BackgroundExploiting Generative AI to Scale up Intelligent Tutoring Systems
Jakubův, J., Chvalovský, K., Goertzel, Z., Kaliszyk, C., Olšák, M., Piotrowski, B., Schulz, S., Suda, M., & Urban, J. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.itp.2023.19 · Cited by 76,658
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173 · cited 4×
“Second, though it is possible to use padding to respect the sentence or other semantic boundaries, in practice it has been standard practice to simply chunk long text into fixed-length segments due to improved efficiency (Peters et al., 2018; Devlin et al., 2018; Al-Rfou et al., 2018).”
BackgroundInfluentialBERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325 · cited 4×
“Second, though it is possible to use padding to respect the sentence or other semantic boundaries, in practice it has been standard practice to simply chunk long text into fixed-length segments due to improved efficiency (Peters et al., 2018; Devlin et al., 2018; Al-Rfou et al., 2018).”
BackgroundInfluentialNeural Machine Translation by Jointly Learning to Align and Translate
Bahdanau, D., Cho, K., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1409.0473 · Cited by 16,213 · cited 2×
“On the other hand, the direct connections between long-distance word pairs baked in attention mechanisms might ease optimization and enable the learning of long-term dependency (Bahdanau et al., 2014; Vaswani et al., 2017).”
BackgroundDeep Contextualized Word Representations
Peters, M., Neumann, M., Iyyer, M., Gardner, M., Clark, C., Lee, K., & Zettlemoyer, L. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers) · doi:10.18653/v1/n18-1202 · Cited by 11,983 · cited 4×
“Second, though it is possible to use padding to respect the sentence or other semantic boundaries, in practice it has been standard practice to simply chunk long text into fixed-length segments due to improved efficiency (Peters et al., 2018; Devlin et al., 2018; Al-Rfou et al., 2018).”
BackgroundInfluentialRecurrent neural network based language model
Mikolov, T., Karafiát, M., Burget, L., Černocký, J., & Khudanpur, S. (2010) · Interspeech 2010 · doi:10.21437/interspeech.2010-343 · Cited by 5,390 · cited 5×
“However, different from truncated BPTT, our method caches a sequence of hidden states instead of the last one, and should be applied together with the relative positional encoding technique described in Section 3.3.”
BackgroundMethodInfluentialAn Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling
Bai, S., Kolter, J. Z., & Koltun, V. (2018) · arXiv · doi:10.48550/arxiv.1803.01271 · Cited by 4,655
Neural Architecture Search with Reinforcement Learning
Zoph, B., & Le, Q. V. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.01578 · Cited by 3,974
Self-citationA Neural Probabilistic Language Model
Bengio, Y., Ducharme, R., Vincent, P., Jauvin, C., Jaz, K., Hofmann, T., Poggio, T., & Shawe-Taylor, J. S. (2000) · Cited by 3,479
DARTS: Differentiable architecture search
Liu, H., Simonyan, K., & Yang, Y. (2018) · arXiv · doi:10.48550/arxiv.1806.09055 · Cited by 3,248
Self-citationGenerating Sequences With Recurrent Neural Networks
Graves, A. (2013) · arXiv · doi:10.48550/arxiv.1308.0850 · Cited by 2,894 · cited 2×
“, 2001), and the introduction of gating in LSTMs and the gradient clipping technique (Graves, 2013) might not be sufficient to fully address this issue.”
BackgroundMethodSelf-Attention with Relative Position Representations
Shaw, P., Uszkoreit, J., & Vaswani, A. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers) · doi:10.18653/v1/n18-2074 · Cited by 2,505 · cited 3×
“Moreover, our relative positional encodings is also superior to Shaw et al. (2018) on short sequences.”
Background(2018) · International Conference on Machine Learning · Cited by 1,465
Self-citationGradient Flow in Recurrent Nets: The Difficulty of Learning LongTerm Dependencies
(2009) · A Field Guide to Dynamical Recurrent Networks · doi:10.1109/9780470544037.ch14 · Cited by 1,246
A theoretically grounded application of dropout in recurrent neural networks
Gal, Y., & Ghahramani, Z. (2016) · Cambridge University Engineering Department Publications Database · Cited by 1,089
Language Modeling with Gated Convolutional Networks
Dauphin, Y., Fan, A., Auli, M., & Grangier, D. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1612.08083 · Cited by 1,080 · cited 1×
“…architectures to better encode the context (Bengio et al., 2003; Mikolov et al., 2010; Zilly et al., 2016; Krause et al., 2016; Grave et al., 2016b; Dauphin et al., 2016; Chung et al., 2016; Merity et al., 2016; Kalchbrenner et al., 2016; Al-Rfou et al., 2018), improving regularization and…”
BackgroundGraves, A., Wayne, G., & Danihelka, I. (2014) · arXiv · doi:10.48550/arxiv.1410.5401 · Cited by 1,058 · cited 1×
“Thus, we can cache a predefined length-M old hidden states spanning (possibly) multiple segments, and refer to them as the memory m nτ ∈ R M × d , due to a clear connection to the memory augmented neural networks (Graves et al., 2014; Weston et al., 2014).”
BackgroundOne Billion Word Benchmark for Measuring Progress in Statistical Language Modeling
Chelba, C., Mikolov, T., Schuster, M., Ge, Q., Brants, T., Koehn, P., & Robinson, T. (2013) · arXiv · doi:10.48550/arxiv.1312.3005 · Cited by 1,051 · cited 5×
“We apply Transformer-XL to a variety of datasets on both word-level and character-level language modeling to have a comparison with state-of-the-art systems, including WikiText-103 (Merity et al., 2016), enwiki8 (LLC, 2009), text8 (LLC, 2009), One Billion Word (Chelba et al., 2013), and Penn Treebank (Mikolov & Zweig,…”
BackgroundInfluentialSemi-supervised Sequence Learning
Dai, A. M., & Le, Q. V. (2015) · arXiv · doi:10.48550/arxiv.1511.01432 · Cited by 994 · cited 2×
“Language modeling is among the important problems that require modeling long-term dependency, with successful applications such as unsupervised pretraining (Dai and Le, 2015; Peters et al., 2018; Radford et al., 2018; Devlin et al., 2018).”
BackgroundSelf-citationIndependently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN
Li, S., Li, W., Cook, C., Zhu, C., & Gao, Y. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00572 · Cited by 972
Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
Shazeer, N., Mirhoseini, A., Maziarz, K., Davis, A., Le, Q., Hinton, G., & Dean, J. (2017) · arXiv · doi:10.48550/arxiv.1701.06538 · Cited by 933
Self-citationExploring the Limits of Language Modeling
Józefowicz, R., Vinyals, O., Schuster, M., Shazeer, N., & Wu, Y. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1602.02410 · Cited by 823 · cited 1×
“…al. (2016); Press & Wolf (2016); Merity et al. (2017); Gal & Ghahramani (2016), speeding up the Softmax computation (Morin & Bengio, 2005; Kuchaiev & Ginsburg, 2017; Grave et al., 2016a; Jozefowicz et al., 2016), and enriching the output distribution family (Yang et al., 2017; Kanai et al., 2018).”
BackgroundA Theoretically Grounded Application of Dropout in Recurrent Neural Networks
Gal, Y., & Ghahramani, Z. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1512.05287 · Cited by 766 · cited 2×
“, 2018), improving regularization and optimization algorithms (Gal and Ghahramani, 2016) , speeding up the Softmax computation (Grave et al.”
Background(2017) · International Conference on Machine Learning · Cited by 758
Regularizing and Optimizing LSTM Language Models
Merity, S., Keskar, N. S., & Socher, R. (2017) · arXiv · doi:10.48550/arxiv.1708.02182 · Cited by 697 · cited 2×
“…al., 2016; Al-Rfou et al., 2018), improving regularization and optimization algorithms Zaremba et al. (2014); Inan et al. (2016); Press & Wolf (2016); Merity et al. (2017); Gal & Ghahramani (2016), speeding up the Softmax computation (Morin & Bengio, 2005; Kuchaiev & Ginsburg, 2017; Grave et al.,…”
BackgroundMethodOne billion word benchmark for measuring progress in statistical language modeling
Chelba, C., Mikolov, T., Schuster, M., Ge, Q., Brants, T., Koehn, P., & Robinson, T. (2014) · Interspeech 2014 · doi:10.21437/interspeech.2014-564 · Cited by 627 · cited 5×
“We apply Transformer-XL to a variety of datasets on both word-level and character-level language modeling to have a comparison with state-of-the-art systems, including WikiText-103 (Merity et al., 2016), enwiki8 (LLC, 2009), text8 (LLC, 2009), One Billion Word (Chelba et al., 2013), and Penn Treebank (Mikolov & Zweig,…”
BackgroundInfluentialTorres H., J. P. (2025) · Zenodo (CERN European Organization for Nuclear Research) · doi:10.5281/zenodo.17754506 · Cited by 580
Context dependent recurrent neural network language model
Mikolov, T., & Zweig, G. (2012) · 2012 IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt.2012.6424228 · Cited by 560 · cited 6×
“We apply Transformer-XL to a variety of datasets on both word-level and character-level language modeling to have a comparison with state-of-the-art systems, including WikiText-103 (Merity et al., 2016), enwiki8 (LLC, 2009), text8 (LLC, 2009), One Billion Word (Chelba et al., 2013), and Penn Treebank (Mikolov & Zweig,…”
BackgroundInfluentialEfficient Neural Architecture Search via Parameter Sharing
Pham, H., Guan, M. Y., Zoph, B., Le, Q. V., & Dean, J. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1802.03268 · Cited by 517
Self-citationA Simple Way to Initialize Recurrent Networks of Rectified Linear Units
Le, Q. V., Jaitly, N., & Hinton, G. E. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1504.00941 · Cited by 513 · cited 2×
“…since the ubiquitous adaption of LSTM, many efforts have been spent on relieving the vanishing gradient problem, including better initialization (Le et al., 2015), additional loss signal (Trinh et al., 2018), augmented memory structure (Ke et al., 2018) and others that modify the internal…”
BackgroundSelf-citationPointer Sentinel Mixture Models
Merity, S., Xiong, C., Bradbury, J., & Socher, R. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1609.07843 · Cited by 427 · cited 10×
“Our best model in this settings use attention lengths of 1,600 and 3,800 on WikiText-103 and enwiki8 respectively.”
BackgroundMethodInfluentialRecurrent Batch NormalizationPreprint
Cooijmans, T., Ballas, N., Laurent, C., Gülçehre, Ç., & Courville, A. C. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1603.09025 · Cited by 424
Neural Machine Translation in Linear Time
Kalchbrenner, N., Espeholt, L., Simonyan, K., Oord, A. V. D., Graves, A., & Kavukcuoglu, K. (2016) · arXiv · doi:10.48550/arxiv.1610.10099 · Cited by 414 · cited 1×
“…Mikolov et al., 2010; Zilly et al., 2016; Krause et al., 2016; Grave et al., 2016b; Dauphin et al., 2016; Chung et al., 2016; Merity et al., 2016; Kalchbrenner et al., 2016; Al-Rfou et al., 2018), improving regularization and optimization algorithms Zaremba et al. (2014); Inan et al. (2016);…”
BackgroundHierarchical Multiscale Recurrent Neural Networks
Chung, J., Ahn, S., & Bengio, Y. (2016) · arXiv · doi:10.48550/arxiv.1609.01704 · Cited by 386 · cited 2×
“In the last few years, the field of language modeling has witnessed many significant advances, including but not limited to devising novel architectures to better encode the context (Bengio et al., 2003; Mikolov et al., 2010; Zilly et al., 2016; Krause et al., 2016; Grave et al., 2016b; Dauphin et al., 2016; Chung et…”
BackgroundSharp nearby, fuzzy far away: How neural language models use context
Khandelwal, U., He, H., Qi, P., & Jurafsky, D. (2018) · Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/p18-1027 · Cited by 303 · cited 1×
“…but not limited to devising novel architectures to better encode the context (Bengio et al., 2003; Mikolov et al., 2010; Zilly et al., 2016; Krause et al., 2016; Grave et al., 2016b; Dauphin et al., 2016; Chung et al., 2016; Merity et al., 2016; Kalchbrenner et al., 2016; Al-Rfou et al.,…”
BackgroundSharp Nearby, Fuzzy Far Away: How Neural Language Models Use Context
Khandelwal, U., He, H., Qi, P., & Jurafsky, D. (2018) · arXiv · doi:10.48550/arxiv.1805.04623 · Cited by 302 · cited 1×
“…but not limited to devising novel architectures to better encode the context (Bengio et al., 2003; Mikolov et al., 2010; Zilly et al., 2016; Krause et al., 2016; Grave et al., 2016b; Dauphin et al., 2016; Chung et al., 2016; Merity et al., 2016; Kalchbrenner et al., 2016; Al-Rfou et al.,…”
BackgroundTying Word Vectors and Word Classifiers: A Loss Framework for Language Modeling
Inan, H., Khosravi, K., & Socher, R. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.01462 · Cited by 288 · cited 1×
“…Merity et al., 2016; Kalchbrenner et al., 2016; Al-Rfou et al., 2018), improving regularization and optimization algorithms Zaremba et al. (2014); Inan et al. (2016); Press & Wolf (2016); Merity et al. (2017); Gal & Ghahramani (2016), speeding up the Softmax computation (Morin & Bengio, 2005;…”
BackgroundEfficient softmax approximation for GPUs
Grave, É., Joulin, A., Cissé, M., & Jeǵou, H. (2017) · International Conference on Machine Learning · Cited by 219
TopicRNN: A Recurrent Neural Network with Long-Range Semantic Dependency
Dieng, A. B., Wang, C., Gao, J., & Paisley, J. W. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.01702 · Cited by 175 · cited 2×
“, 2015; Wang and Cho, 2015) to others that rely on document-level topics learned from data (Dieng et al., 2016; Wang et al., 2017).”
BackgroundEfficient softmax approximation for GPUs
Grave, É., Joulin, A., Cissé, M., Grangier, D., & Jeǵou, H. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1609.04309 · Cited by 132 · cited 2×
“We adopted adaptive softmax and input representations (Baevski and Auli, 2018; Grave et al., 2016a).”
BackgroundMethodAn Analysis of Neural Language Modeling at Multiple Scales
Merity, S., Keskar, N. S., & Socher, R. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.08240 · Cited by 123 · cited 2×
“The “Transformer-XL (151M)” setting uses a standard parameter budget as previous work Merity et al. (2018), where we observe a similar effect when increasing the attention length during evaluation.”
MethodHa, D., Dai, A., & Le, Q. V. (2016) · arXiv · doi:10.48550/arxiv.1609.09106 · Cited by 100
Self-citationFactorization tricks for LSTM networks
Kuchaiev, O., & Ginsburg, B. (2017) · arXiv · doi:10.48550/arxiv.1703.10722 · Cited by 99 · cited 1×
“…Inan et al. (2016); Press & Wolf (2016); Merity et al. (2017); Gal & Ghahramani (2016), speeding up the Softmax computation (Morin & Bengio, 2005; Kuchaiev & Ginsburg, 2017; Grave et al., 2016a; Jozefowicz et al., 2016), and enriching the output distribution family (Yang et al., 2017; Kanai et…”
BackgroundLearning Longer-term Dependencies in RNNs with Auxiliary Losses
Trinh, T. H., Dai, A. M., Luong, M.-T., & Le, Q. V. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.00144 · Cited by 84 · cited 2×
“…many efforts have been spent on relieving the vanishing gradient problem, including better initialization (Le et al., 2015), additional loss signal (Trinh et al., 2018), augmented memory structure (Ke et al., 2018) and others that modify the internal architecture of RNNs to ease the optimization…”
BackgroundSelf-citationImproving Neural Language Models with a Continuous Cache
Grave, É., Joulin, A., & Usunier, N. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1612.04426 · Cited by 83 · cited 3×
“We adopted adaptive softmax and input representations (Baevski & Auli, 2018; Grave et al., 2016a).”
BackgroundMethodMultiplicative LSTM for sequence modelling
Krause, B., Lu, L., Murray, I., & Renals, S. J. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1609.07959 · Cited by 78 · cited 2×
“…including but not limited to devising novel architectures to better encode the context (Bengio et al., 2003; Mikolov et al., 2010; Zilly et al., 2016; Krause et al., 2016; Grave et al., 2016b; Dauphin et al., 2016; Chung et al., 2016; Merity et al., 2016; Kalchbrenner et al., 2016; Al-Rfou et al.,…”
BackgroundZilly, J., Srivastava, R., Koutník, J., & Schmidhuber, J. (2016) · arXiv (Cornell University) · doi:10.3929/ethz-b-000233981 · Cited by 74
An Improved Relative Self-Attention Mechanism for Transformer with Application to Music GenerationPreprint
Huang, C.-Z. A., Vaswani, A., Uszkoreit, J., Shazeer, N., Hawthorne, C., Dai, A. M., Hoffman, M. D., & Eck, D. (2018) · arXiv (Cornell University) · Cited by 73
Breaking the Softmax Bottleneck: A High-Rank RNN Language Model
Yang, Z., Dai, Z., Salakhutdinov, R., & Cohen, W. W. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1711.03953 · Cited by 62 · cited 2×
“, 2016a) , and enriching the output distribution family (Yang et al., 2017).”
BackgroundSelf-citationSparse Attentive Backtracking: Temporal CreditAssignment Through Reminding
Ke, N. R., Goyal, A., Bilaniuk, O., Binas, J., Mozer, M. C., Pal, C., & Bengio, Y. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.03702 · Cited by 58
Fast-Slow Recurrent Neural Networks
Mujika, A., Meier, F., & Steger, A. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1705.08639 · Cited by 57
Character-Level Language Modeling with Deeper Self-Attention
Al-Rfou, R., Choe, D., Constant, N., Guo, M., & Jones, L. (2019) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v33i01.33013159 · Cited by 56
Document Context Language Models
Ji, Y., Cohn, T., Kong, L., Dyer, C., & Eisenstein, J. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1511.03962 · Cited by 55 · cited 2×
“ually defined (Mikolov and Zweig, 2012; Ji et al., 2015; Wang and Cho, 2015) to others that rely on document-level topics learned from data (Dieng et al.”
BackgroundOn Multiplicative Integration with Recurrent Neural Networks
Wu, Y., Zhang, S., Zhang, Y., Bengio, Y., & Salakhutdinov, R. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.06630 · Cited by 54 · cited 1×
“…better initialization (Le et al., 2015), additional loss signal (Trinh et al., 2018), augmented memory structure (Ke et al., 2018) and others that modify the internal architecture of RNNs to ease the optimization Mikolov et al. (2014); Koutnik et al. (2014); Wu et al. (2016); Li et al. (2018).”
BackgroundSelf-citationMesh-TensorFlow: Deep Learning for Supercomputers
Shazeer, N., Cheng, Y., Parmar, N., Tran, D., Vaswani, A., Koanantakool, P., Hawkins, P., Lee, H., Hong, M., Young, C., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.02084 · Cited by 50
Topic Compositional Neural Language Model
Wang, W., Gan, Z., Wang, W., Shen, D., Huang, J., Ping, W., Satheesh, S., & Carin, L. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1712.09783 · Cited by 49 · cited 2×
“, 2015; Wang and Cho, 2015) to others that rely on document-level topics learned from data (Dieng et al., 2016; Wang et al., 2017).”
BackgroundAdaptive Input Representations for Neural Language Modeling
Baevski, A., & Auli, M. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.10853 · Cited by 48 · cited 4×
“We adopted adaptive softmax and input representations (Baevski and Auli, 2018; Grave et al., 2016a).”
MethodInfluentialLarger-Context Language Modelling with Recurrent Neural Network
Wang, T., & Cho, K. (2016) · Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/p16-1125 · Cited by 47 · cited 1×
“Existing works range from ones where context representations are manually defined (Mikolov & Zweig, 2012; Ji et al., 2015; Wang & Cho, 2015) to others that rely on document-level topics learned from data (Dieng et al., 2016; Wang et al., 2017).”
BackgroundLarger-Context Language Modelling
Wang, T., & Cho, K. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1511.03729 · Cited by 38 · cited 1×
“Existing works range from ones where context representations are manually defined (Mikolov & Zweig, 2012; Ji et al., 2015; Wang & Cho, 2015) to others that rely on document-level topics learned from data (Dieng et al., 2016; Wang et al., 2017).”
BackgroundFast Parametric Learning with Activation Memorization
Rae, J. W., Dyer, C., Dayan, P., & Lillicrap, T. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.10049 · Cited by 20
Skip-gram Language Modeling Using Sparse Non-negative Matrix Probability EstimationPreprint
Shazeer, N., Pelemans, J., & Chelba, C. I. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1412.1454 · Cited by 13
Pushing the bounds of dropoutPreprint
Melis, G., Blundell, C., Kočiský, T., Hermann, K. M., Dyer, C., & Blunsom, P. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.09208 · Cited by 10
Weston, J., Chopra, S., & Bordes, A. (2014) · arXiv · doi:10.48550/arxiv.1410.3916 · cited 1×
“Thus, we can cache a predefined length-M old hidden states spanning (possibly) multiple segments, and refer to them as the memory m nτ ∈ R M × d , due to a clear connection to the memory augmented neural networks (Graves et al., 2014; Weston et al., 2014).”
Background
Research works printed but not linked: 1
- Matthew E Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. 2018. Deep contextualized word rep- resentations. arXiv preprint arXiv:1802.05365. · arXiv
Cited by
Transformers: State-of-the-Art Natural Language Processing
Wolf, T., Debut, L., Sanh, V., Chaumond, J., Delangue, C., Moi, A., Cistac, P., Rault, T., Louf, R., Funtowicz, M., et al. (2020) · Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations · doi:10.18653/v1/2020.emnlp-demos.6 · Cited by 8,741
“, 2019) Trans-XL (Dai et al., 2019) XLNet (Yang et al.”
Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
Zhou, H., Zhang, S., Peng, J., Zhang, S., Li, J., Xiong, H., & Zhang, W. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i12.17325 · Cited by 7,752
Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers
Zheng, S., Lu, J., Zhao, H., Zhu, X., Luo, Z., Wang, Y., Fu, Y., Feng, J., Xiang, T., Torr, P. H. S., et al. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00681 · Cited by 4,008
Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers
Zheng, S., Lu, J., Zhao, H., Zhu, X., Luo, Z., Wang, Y., Fu, Y., Feng, J., Xiang, T., Torr, P. H. S., et al. (2020) · arXiv · doi:10.48550/arxiv.2012.15840 · Cited by 4,004
A review on the attention mechanism of deep learning
Niu, Z., Zhong, G., & Yu, H. (2021) · Neurocomputing · doi:10.1016/j.neucom.2021.03.091 · Cited by 3,379
Metadata
- Authors
- Dai, Z., Yang, Z., Yang, Y., Carbonell, J., Le, Q., & Salakhutdinov, R.
- Affiliations
- Carnegie Mellon University (US)
- Google (United States) (US)
- Venue
- Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics
- Published
- 2019
- Pages
- 2978-2988
- ISSN
- 0736-587X (ISSN-L)
- OpenAlex
- W2964110616
- Cited by
- 3,169
- References
- 67
- Citations by year
- 2026: 163 · 2025: 361 · 2024: 434 · 2023: 585 · 2022: 504 · 2021: 612 · 2020: 403 · 2019: 107
- FWCI
- 155.55
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Proceedings of the conference - Association for Computational Linguistics. Meeting (publishedVersion, cc-by)
- Funders
- National Science Foundation (1546329, 1763562)
- U.S. Department of Energy
- Nvidia
- Office of Science
- Office of Naval Research (N000141812861)
- Advanced Scientific Computing Research
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Keywords
- Transformer-XL, long-range dependencies, positional encoding, language modeling, temporal coherence, sequence generation, pre-trained language models
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.

