Sarcastic Headline Generator
Can a character-level RNN learn sarcasm? I trained three architectures on The Onion to find out, comparing vanilla RNN, LSTM, and GRU performance on context-dependent irony.
resources
- report.pdf (paper)
- code (code)
Can a character-level RNN learn sarcasm? I trained three architectures on The Onion to find out, comparing vanilla RNN, LSTM, and GRU performance on context-dependent irony.