Guided Transformer: Updating Transformer Representations through Augmentation

◾Guided Transformer: Updating Transformer Representations through
Retrieval Augmentation

◾ Speaker:
Helia Hashemi

◾ Talk Description:
This talk is part of the #women_in_nlp talk series, inviting women who successfully carved their career path in NLP to share their experiences and advice. Everyone is welcome to attend the talk not only women.

◾ Abstract:
In this talk, I will introduce Guided Transformer, a generic extension
to Transformer networks that adjusts the input representation of a
network using cross attention from multiple information items. These
information items can be retrieved from a large corpus. Thus, Guided
Transformer has numerous applications in retrieval-augmented neural
networks. I will present its applications to conversational search and
query representation learning tasks and discuss potential future
directions.

◾ Bio:
Helia Hashemi is a fifth-year Ph.D. student at the Center for
Intelligent Information Retrieval (CIIR) of the University of
Massachusetts Amherst. She is a recipient of Bloomberg's Data Science
Ph.D. Fellowship. Her research interests lie in the areas of
Information Retrieval and Natural Language Processing. In her Ph.D., she
has focused on conversational information-seeking systems, or in other
words, how to help people find the information they need through
conversation -- the most natural way of communication. She did research internships at Spotify and Google Research. Helia's website is https://people.cs.umass.edu/~hhashemi/

◾ About #women_in_nlp
Website: https://efatmae.github.io/womenInNlp/
Twitter: https://twitter.com/fatmaElsafoury
Slack channel on dair.ai: https://dairai.slack.com/archives/C01J0GXJMD1

◾ About dair.ai
Website: https://dair.ai/
GitHub: https://github.com/dair-ai
Twitter: https://twitter.com/dair_ai
Discord: https://discord.gg/SKgkVT8BGJ
Slack: https://join.slack.com/t/dairai/shared_invite/zt-pcxkmoip-b4nJkci8L_dynpMwLvlCcQ

◾ Code of Conduct: https://github.com/dair-ai/dair-ai.github.io/blob/master/CODE_OF_CONDUCT.md