Papers › Densely Connected Attention Propagation for Reading Comprehension

Densely Connected Attention Propagation for Reading Comprehension

10 Nov 2018NeurIPS 2018 12arXiv:1811.04210archive 2025-07-28

Yi Tay, Luu Anh Tuan, Siu Cheung Hui, Jian Su

We propose DecaProp (Densely Connected Attention Propagation), a new densely connected neural architecture for reading comprehension (RC). There are two distinct characteristics of our model. Firstly, our model densely connects all pairwise layers of the network, modeling relationships between passage and query across all hierarchical levels. Secondly, the dense connectors in our network are learned via attention instead of standard residual skip-connectors. To this end, we propose novel Bidirectional Attention Connectors (BAC) for efficiently forging connections throughout the network. We conduct extensive experiments on four challenging RC benchmarks. Our proposed approach achieves state-of-the-art results on all four, outperforming existing baselines by up to 2.6%-14.2% in absolute F1 score.

PaperPDFConference PDFCode

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

Code

vanzytay/NIPS2018_DECAPROP mentioned on GitHubtf 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

AllOpen-Domain Question AnsweringQuestion AnsweringReading Comprehension

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Open-Domain Question Answering Quasar DecaProp EM (Quasar-T) 38.6 #3 of 6 Archive leaderboard report
Open-Domain Question Answering Quasar DecaProp F1 (Quasar-T) 46.9 #3 of 6 Archive leaderboard report
Open-Domain Question Answering SearchQA DECAPROP EM 62.2 #5 of 14 Archive leaderboard report
Open-Domain Question Answering SearchQA DecaProp EM 56.8 #7 of 14 Archive leaderboard report
Open-Domain Question Answering SearchQA DecaProp F1 63.6 #7 of 14 Archive leaderboard report
Open-Domain Question Answering SearchQA DecaProp N-gram F1 70.8 #7 of 14 Archive leaderboard report
Open-Domain Question Answering SearchQA DecaProp Unigram Acc 62.2 #7 of 14 Archive leaderboard report
Question Answering NarrativeQA DecaProp BLEU-1 44.35 #5 of 10 Archive leaderboard report
Question Answering NarrativeQA DecaProp BLEU-4 27.61 #5 of 10 Archive leaderboard report
Question Answering NarrativeQA DecaProp METEOR 21.80 #5 of 10 Archive leaderboard report
Question Answering NarrativeQA DecaProp Rouge-L 44.69 #5 of 10 Archive leaderboard report
Question Answering NewsQA DecaProp EM 53.1 #12 of 18 Archive leaderboard report
Question Answering NewsQA DecaProp F1 66.3 #12 of 18 Archive leaderboard report
Question Answering Quasart-T DECAPROP EM 38.6 #6 of 7 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