Reviews: Latent Attention For If-Then Program Synthesis

But Table 2. shows BDLSTM (average 10 models) improves over [14]. Please also check the same in the line nos.