3 accepted papers
We propose a confidence scoring mechanism for multi-layer neural networks based on a paradigm of a base model and a meta-model. The confidence score is learned by the meta-model using features derived from the base model – a deep neural network considered a whitebox. As features, we investigate line…
In this paper, we study the problem of visual reasoning in the context of textual question answering. We introduce Dynamic Spatial Memory Networks (DSMN), a new deep network architecture that specializes in answering questions that admit latent visual representations, and learns to generate and reas…
Learning effective text representations is a key foundation for numerous machine learning and NLP applications. While the celebrated Word2Vec technique yields semantically rich word representations, it is less clear whether sentence or document representations should be built upon word representatio…