Internship Offer at INRIA Title: Building test cases for understanding large neural-network based dialog models Description: Conversational Systems have made huge progress in the past years. Agents like Siri, Amazon Alexa and Google Assistants have helped bring such state-of-the-art conversational agents into the daily lives of millions of people. Large artificial neural-network based dialog models have played a huge role in making the machines talk like humans. Some of these models include Facebook's BlenderBot [1] and Microsoft's DialoGPT[2]. They have been able to produce grammatically correct responses to the previous utterances from the users. However, these models are far from having the ability to understand and converse like humans in natural language. While, they seem to have obtained the syntactic and to some extent the semantic knowledge of the conversations, they still lack in their ability to understand the intended meaning of an utterance uttered by humans. This, could be seen as the lack of understanding of the pragmatics by these models often leading to inconsistent responses over a relatively long period of conversation. In order to solve this problem, we need to understand what these large dialog models learn in the first place. The insights from this study would later help us develop algorithms to solve the larger problem. Hence, we are offering an internship where the intern would work closely with a PhD student in order to build various test cases to study the dialog models. The internship requires good background in linguistics(high priority). Decent coding knowledge is desirable but not necessary. These projects take place in the ArticuLabo at Inria Paris Center (with a joint affiliation with Carnegie Mellon University in the US) headed by Professor Justine Cassell. To apply, please send a CV and a letter of motivation to biswesh.mohapatra@inria.fr About Inria Inria is the French national research institute dedicated to digital science and technology. It employs 2,600 people. Its 200 agile project teams, generally run jointly with academic partners, include more than 3,500 scientists and engineers working to meet the challenges of digital technology, at the interface with other disciplines.The Institute also employs numerous talents in over forty different professions. 900 research support staff contribute to the preparation and development of scientific and entrepreneurial projects that have a worldwide impact. References: [1] Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Eric Michael Smith, Y-Lan Boureau, and Jason Weston. 2021. Recipes for Building an Open-Domain Chatbot. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 300-325, Online. Association for Computational Linguistics. [2] Zhang, Yizhe, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu and William B. Dolan. "DIALOGPT : Large-Scale Generative Pre-training for Conversational Response Generation." ACL (2020).