Page history
14 March 2021
→2016
+128
→2017
+100
→2018
+30
→References
+21,531
→Soft and Hard Attention
+37
ContinuousReplacement
+1
→References
+495
→9. Conclusion
+714
→9. Conclusion
+1,352
→8. Related Work
+8,066
ContinuousReplacement
+2
→7. Limitations of Our Model
+4,222
→6.4.2 Semantic Properties of the Learned Embeddings
+2
→6.4.2 Semantic Properties of the Learned Embeddings
−64
→Path-Attention Interpretation
−1
→6.4.2 Semantic Properties of the Learned Embeddings
+5,628
→6.4.1 Interpreting Attention
+4,149
→6. Evaluation
−38
→1. Introduction
−26
→4. Model
−6
→5. Distributed Vs. Symbolic Representations
−18
→4.5 Design Decisions
−27
→= Predicting Tags and Names
+1
→4.4 Using the Trained Network
−15
→4.2 Path-Attention Model
+4
→4.1 Code as a Bag of Path-Contexts
+1
→2.1 Motivating Example
−8
ContinuousReplacement
+2
→1.2 Challenges: Representation and Attention
−10
→6.4 Qualitative Evaluation
+105
→6.3 Data Ablation Study
+3,138
→4.2 Path-Attention Model
→6.2 Evaluation of Alternative Designs
+4,983
→6.1 Quantitative Evaluation
+7,630
13 March 2021
ContinuousReplacement
+4
→6. Evaluation
+1,681
→6. Evaluation
+4,725
→5. Distributed Vs. Symbolic Representations
+4,859
→4.5 Design Decisions
+8,396
→4.4 Using the Trained Network
+2,212
→4.2 Path-Attention Model
+157
→4.3 Training
+1,579
8 March 2021
7 March 2021
→4.2 Path-Attention Model
+7,090
→4.2 Path-Attention Model
+2,452
→4.1 Code as a Bag of Path-Contexts
+1,791
ContinuousReplacement
+4
→4. Model
+2,562