Research Scientist
Infolab @ Mit
Cambridge, Massachusetts, United States
"Zero-shot linear combinations of grounded social interactions with Linear Social MDPs." AAAI 2023-Implemented a recursive MDP (Markov Decision Processes) framework for incorporating social interactions in multi-agent settings using dynamic programming.-Specified reward functions in terms of the estimated reward functions of other agents, modeling social interactions such as helping or hindering another agent (by maximizing or minimizing the other agent reward)."Incorporating Rich Social Interactions Into MDPs." IEEE, 2022 -Developed a python interface for simulating multi-agent interaction.-Led experiments for data collection."The Aligned Multimodal Movie Treebank: An audio, video, dependency-parse treebank." EMNLP, 2022 -Annotated electrocorticogram (ECOG) data, which traces brain waves representing dialog transcripts from Hollywood movies. Defined part of speech tags and dependency parses.-The work consisted of 31,264 sentences and 218,090 words amounting to the 3rd largest Universal Dependencies (UD) English treebank.