Workshop @ EMNLP 2021
Modelling language based human-agent interactions has been an important topic of study in natural language processing [1, 2, 3]. Recent state-of-the-art models can have task-driven conversations or engage users in an open-dialogue on a given topic. Efforts on extending the human-agent interaction for advising reinforcement learning agents with text-based rewards can be seen as a promising first step towards using language as a medium of instruction to a task-learning agent . Further research in cross-modal learning like VisualQA [5, 6, 7] and Embodied learning [8, 9, 10] have enabled learning agents to make use of the information available across modalities to solve a task through interactions.
With recent promise shown by the research in using language as a mode of instruction for learning agents [11, 12, 13, 14] there is, now, scope for accelerating the discussions on realizing domains and discovering novel problems in the topic of Learning-to-Learn through interaction. Progress along this research direction can potentially facilitate agents to have advanced interactions like asking for more samples, defining its learning limits, expanding its capacity, or regularizing its learning and even defining a new task which can be used towards building effective learning-to-learn mechanisms. NILLI@EMNLP will be focusing the discussions on continual interact-and-learn procedure to systematically acquire knowledge and solve tasks through verbal and non-verbal interactions.
We invite you to take part in NILLI@EMNLP21 that fosters discussions on nurturing an interdisciplinary research unifying the paradigms of continual learning, natural language processing, embodied learning, reinforcement learning, robot learning and multi-modal learning towards building interactive and interpretable AI.
Associate Professor, University of Washington
Research Scientist, Deepmind
Research Scientist, Google Brain
Research Scientist, Facebook AI Research
Professor, Carnegie Mellon University
We call for novel, unpublished or in-review works on topics listed in the groups below:
- Language at Fore
- Novel environments for language understanding through interaction.
- Language based Reinforcement Learning.
- Learning representations in grounded language.
- Language based interaction methods in interdisciplinary research.
- Machine Learning with Interaction
- Modeling multi-modal and language interactions to aid continual learning.
- Interactive training for embodied agents.
- Early and negative results on learning to solve tasks through interaction.
- Non-verbal/Verbal interactive frameworks.
- Community Impact of Interactive Agents
- Frontiers in building interactive agents (data, frameworks, open problems).
- Applications of interactive learning in interdisciplinary research.
- Security and ethical challenges in interaction based learning.
- Submission deadline:
August 30September 10, 2021
- Notification of acceptance: September 29, 2021
- Camera-ready papers due: October 10, 2021
- Workshop: November
All deadlines are 11.59 pm UTC -12h (“anywhere on Earth”).
The submission link is open at CMT. The style guidelines follows the format guidelines of EMNLP 2021.
Authors of any papers that are submitted to ARR before the NILLI’s submission deadline, and that have reviews by September 23 may submit their papers and reviews for NILLI.
There is a weekly office hour on Fridays@11AM EST over zoom. (Note: The calls will not be recorded and the camera will be off by default.)
Please book an appointment. As the office hour is run on-demand.
Senior Ph.D. Candidate,
Senior Ph.D. Candidate
Polytechnique de Montreal.
- Nikita Moghe, University of Edinburgh
- Janarthanan Rajendran, University of Michigan
- Eric Yuan, MSR Montreal
- Danni Ma, University of Pennsylvania
- Richard Pang, New York University
- Sneha Mehta, Virginia Tech
- Sashank Santhanam, University of North Carolina
- Luciana Benotti, Universidad Nacional de Córdoba
- Roma Patel, Brown University
- Bishal Santra, IIT Kharagpur
- Ifrah Idrees, Brown University
- Siddharth Karamcheti, Stanford University
- Thao Nguyen, Brown University
- Yu-Siang Wang, University of Toronto
- Yow-Ting Shiue, University of Maryland
- Josh Roy, Brown University
- Eric Rosen, Brown University
- Ta-Chung Chi, Carnegie Mellon University