Papers › Direct Output Connection for a High-Rank Language Model

Direct Output Connection for a High-Rank Language Model

30 Aug 2018EMNLP 2018 10arXiv:1808.10143archive 2025-07-28

Sho Takase, Jun Suzuki, Masaaki Nagata

This paper proposes a state-of-the-art recurrent neural network (RNN) language model that combines probability distributions computed not only from a final RNN layer but also from middle layers. Our proposed method raises the expressive power of a language model based on the matrix factorization interpretation of language modeling introduced by Yang et al. (2018). The proposed method improves the current state-of-the-art language model and achieves the best score on the Penn Treebank and WikiText-2, which are the standard benchmark datasets. Moreover, we indicate our proposed method contributes to two application tasks: machine translation and headline generation. Our code is publicly available at: https://github.com/nttcslab-nlp/doc_lm.

PaperPDFConference PDFCode

In Syntology View this paper on Syntology: its repositories, every harvested function with whether it ran, its licence and the call to fetch it.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

nttcslab-nlp/doc_lm officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Constituency ParsingHeadline GenerationLanguage ModelingLanguage ModellingMachine TranslationTranslationVocal Bursts Intensity Prediction

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Constituency Parsing Penn Treebank LSTM Encoder-Decoder + LSTM-LM F1 score 94.47 #17 of 27 Archive leaderboard report
Language Modelling Penn Treebank (Word Level) AWD-LSTM-DOC x5 Params 185M #8 of 43 Archive leaderboard report
Language Modelling Penn Treebank (Word Level) AWD-LSTM-DOC x5 Test perplexity 47.17 #8 of 43 Archive leaderboard report
Language Modelling Penn Treebank (Word Level) AWD-LSTM-DOC x5 Validation perplexity 48.63 #8 of 43 Archive leaderboard report
Language Modelling Penn Treebank (Word Level) AWD-LSTM-DOC Params 23M #16 of 43 Archive leaderboard report
Language Modelling Penn Treebank (Word Level) AWD-LSTM-DOC Test perplexity 52.38 #16 of 43 Archive leaderboard report
Language Modelling Penn Treebank (Word Level) AWD-LSTM-DOC Validation perplexity 54.12 #16 of 43 Archive leaderboard report
Language Modelling WikiText-2 AWD-LSTM-DOC x5 Number of params 185M #20 of 38 Archive leaderboard report
Language Modelling WikiText-2 AWD-LSTM-DOC x5 Test perplexity 53.09 #20 of 38 Archive leaderboard report
Language Modelling WikiText-2 AWD-LSTM-DOC x5 Validation perplexity 54.19 #20 of 38 Archive leaderboard report
Language Modelling WikiText-2 AWD-LSTM-DOC Number of params 37M #24 of 38 Archive leaderboard report
Language Modelling WikiText-2 AWD-LSTM-DOC Test perplexity 58.03 #24 of 38 Archive leaderboard report
Language Modelling WikiText-2 AWD-LSTM-DOC Validation perplexity 60.29 #24 of 38 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections