Difference between revisions of "2016 ADecomposableAttentionModelforN"

From GM-RKB
Jump to: navigation, search
(Abstract)
(ContinuousReplacement)
(Tag: continuous replacement)
Line 18: Line 18:
 
On the [[Stanford Natural Language Inference (SNLI) dataset]], [[2016_ADecomposableAttentionModelforN|we]] obtain [[state-of-the-art results]] with almost an [[order of magnitude]] [[fewer parameter]]s than [[previous work]] and without relying on any [[word-order information]]. </s>
 
On the [[Stanford Natural Language Inference (SNLI) dataset]], [[2016_ADecomposableAttentionModelforN|we]] obtain [[state-of-the-art results]] with almost an [[order of magnitude]] [[fewer parameter]]s than [[previous work]] and without relying on any [[word-order information]]. </s>
 
Adding [[intra-sentence attention]] that takes a [[minimum amount]] of order into account yields further improvements. </s>
 
Adding [[intra-sentence attention]] that takes a [[minimum amount]] of order into account yields further improvements. </s>
 +
 
=== 1 Introduction ===
 
=== 1 Introduction ===
 +
 
=== 2 Related Work ===
 
=== 2 Related Work ===
 +
 
=== 3 Approach ===
 
=== 3 Approach ===
 +
 
==== 3.1 Attend ====
 
==== 3.1 Attend ====
 +
 
==== 3.2 Compare ====
 
==== 3.2 Compare ====
 +
 
==== 3.3 Aggregate ====
 
==== 3.3 Aggregate ====
 +
 
==== 3.4 Intra-Sentence Attention (Optional) ====
 
==== 3.4 Intra-Sentence Attention (Optional) ====
 +
 
=== 4 Computational Complexity ===
 
=== 4 Computational Complexity ===
 +
 
=== 5 Experiments ===
 
=== 5 Experiments ===
 +
 
==== 5.1 Implementation Details ====
 
==== 5.1 Implementation Details ====
 +
 
==== 5.2 Results ====
 
==== 5.2 Results ====
 +
 
=== 6 Conclusion ===
 
=== 6 Conclusion ===
 +
 
=== Acknowledgements ===
 
=== Acknowledgements ===
  

Revision as of 00:37, 13 September 2019

Subject Headings: Attention Mechanism, Textual Entailment Recognition.

Notes

Cited By

Quotes

Abstract

We propose a simple neural architecture for natural language inference. Our approach uses attention to decompose the problem into subproblems that can be solved separately, thus making it trivially parallelizable. On the Stanford Natural Language Inference (SNLI) dataset, we obtain state-of-the-art results with almost an order of magnitude fewer parameters than previous work and without relying on any word-order information. Adding intra-sentence attention that takes a minimum amount of order into account yields further improvements.

1 Introduction

2 Related Work

3 Approach

3.1 Attend

3.2 Compare

3.3 Aggregate

3.4 Intra-Sentence Attention (Optional)

4 Computational Complexity

5 Experiments

5.1 Implementation Details

5.2 Results

6 Conclusion

Acknowledgements

References

;

 AuthorvolumeDate ValuetitletypejournaltitleUrldoinoteyear
2016 ADecomposableAttentionModelforNAnkur P. Parikh
Oscar Tackstrom
Dipanjan Das
Jakob Uszkoreit
A Decomposable Attention Model for Natural Language Inference2016
AuthorAnkur P. Parikh +, Oscar Tackstrom +, Dipanjan Das + and Jakob Uszkoreit +
titleA Decomposable Attention Model for Natural Language Inference +
year2016 +