Research & Projects
Research & Projects
Embedded Files
Co-Authorship
Co-Authorship
Benchmarking generalization via in-context instructions on 1,600+ language tasks
Benchmarking generalization via in-context instructions on 1,600+ language tasks
arXiv preprint arXiv:2204.07705
Conducted extensive analysis and evaluation of generalization performance across a diverse range of 100+ language tasks, providing valuable insights into the effectiveness of in-context instructions.
Objectives
Objectives
Helped create 100+ instruction sets that can be used as a benchmark to check the effectiveness of translation of NLP models on language data.
Page updated
Report abuse