Please read at least two papers from the list. For each paper you read, submit brief answers for the following:

  • The title of the paper
  • Problem definition
  • Technical challenges
  • Summary of approach/main ideas
  • Evaluation methods and results
  • Identify limitations. How would you change the approach to improve?
  • Discussion points

Papers

  • Ouyang, L. et. al. “InstructGPT: Training language models to follow instructions with human feedback” (2022).ArXiv
  • Driess, D. et. al “PaLM-E: An Embodied Multimodal Language Model” (2023).ArXiv
  • Wang, G. et. al. “Voyager: An Open-Ended Embodied Agent with Large Language Models.” (2023). ArXiv