Hi, I am Sebastian and I am a Ph.D. candidate in Computer Science at Harvard SEAS. I am working on the following:
- Methods to explore and interact with deep learning models.
- NLG tasks such as summarization and conditional language modeling.
I want to build useful and interpretable models that people can interact with and that support people when making decisions.
My academic advisors are Barbara Grosz and Sasha Rush. I am primarily working with the Harvard NLP Group, and sometimes the Intelligent Interactive Systems Group. For more information about me, you can have a look at my CV (last updated March '19).
Feel free to contact me with requests for discussions, questions about my research or whatever you feel the need of talking to me about. My email-address is my_last_name (at) seas.harvard.edu or you can follow me on twitter @SebGehr.
- End-to-End Content and Plan Selection for Data-to-Text Generation
- Bottom-up abstractive summarization
- Seq2Seq-Vis: A Visual Debugging Tool for Sequence-to-Sequence Models
- End-to-End Content and Plan Selection for Natural Language Generation
- LSTMVis: A Tool for Visual Analysis of Hidden State Dynamics in Recurrent Neural Networks
- Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives