Sanghyun Yi | California Institute of Technology (original) (raw)
Social and Decision Neuroscience
less
Uploads
Papers by Sanghyun Yi
Proceedings of the 12th International Conference on Natural Language Generation
Encoder-decoder based neural architectures serve as the basis of state-of-the-art approaches in e... more Encoder-decoder based neural architectures serve as the basis of state-of-the-art approaches in end-to-end open domain dialog systems. Since most of such systems are trained with a maximum likelihood (MLE) objective they suffer from issues such as lack of generalizability and the generic response problem, i.e., a system response that can be an answer to a large number of user utterances, e.g., "Maybe, I don't know." Having explicit feedback on the relevance and interestingness of a system response at each turn can be a useful signal for mitigating such issues and improving system quality by selecting responses from different approaches. Towards this goal, we present a system that evaluates chatbot responses at each dialog turn for coherence and engagement. Our system provides explicit turn-level dialog quality feedback, which we show to be highly correlated with human evaluation. To show that incorporating this feedback in the neural response generation models improves dialog quality, we present two different and complementary mechanisms to incorporate explicit feedback into a neural response generation model: reranking and direct modification of the loss function during training. Our studies show that a response generation model that incorporates these combined feedback mechanisms produce more engaging and coherent responses in an open-domain spoken dialog setting, significantly improving the response quality using both automatic and human evaluation.
It is an open question whether preferences for visual art can be lawfully predicted from the basi... more It is an open question whether preferences for visual art can be lawfully predicted from the basic constituent elements of a visual image. Moreover, little is known about how such preferences are actually constructed in the brain. Here we developed and tested a computational framework to gain an understanding of how the human brain constructs aesthetic value. We show that it is possible to explain human preferences for a piece of art based on an analysis of features present in the image. This was achieved by analyzing the visual properties of drawings and photographs by multiple means, ranging from image statistics extracted by computer vision tools, subjective human ratings about attributes, to a deep convolutional neural network. Crucially, it is possible to predict subjective value ratings not only within but also across individuals, speaking to the possibility that much of the variance in human visual preference is shared across individuals. Neuroimaging data revealed that prefe...
Proceedings of the 12th International Conference on Natural Language Generation
Encoder-decoder based neural architectures serve as the basis of state-of-the-art approaches in e... more Encoder-decoder based neural architectures serve as the basis of state-of-the-art approaches in end-to-end open domain dialog systems. Since most of such systems are trained with a maximum likelihood (MLE) objective they suffer from issues such as lack of generalizability and the generic response problem, i.e., a system response that can be an answer to a large number of user utterances, e.g., "Maybe, I don't know." Having explicit feedback on the relevance and interestingness of a system response at each turn can be a useful signal for mitigating such issues and improving system quality by selecting responses from different approaches. Towards this goal, we present a system that evaluates chatbot responses at each dialog turn for coherence and engagement. Our system provides explicit turn-level dialog quality feedback, which we show to be highly correlated with human evaluation. To show that incorporating this feedback in the neural response generation models improves dialog quality, we present two different and complementary mechanisms to incorporate explicit feedback into a neural response generation model: reranking and direct modification of the loss function during training. Our studies show that a response generation model that incorporates these combined feedback mechanisms produce more engaging and coherent responses in an open-domain spoken dialog setting, significantly improving the response quality using both automatic and human evaluation.
It is an open question whether preferences for visual art can be lawfully predicted from the basi... more It is an open question whether preferences for visual art can be lawfully predicted from the basic constituent elements of a visual image. Moreover, little is known about how such preferences are actually constructed in the brain. Here we developed and tested a computational framework to gain an understanding of how the human brain constructs aesthetic value. We show that it is possible to explain human preferences for a piece of art based on an analysis of features present in the image. This was achieved by analyzing the visual properties of drawings and photographs by multiple means, ranging from image statistics extracted by computer vision tools, subjective human ratings about attributes, to a deep convolutional neural network. Crucially, it is possible to predict subjective value ratings not only within but also across individuals, speaking to the possibility that much of the variance in human visual preference is shared across individuals. Neuroimaging data revealed that prefe...