The 2022 Reprogen Shared Task On Reproducibility Of Evaluations In NLG: Overview And Results

Anya Belz, Anastasia Shimorina, Maja Popovic, Ehud Reiter

GenChal - Thursday 07/21 11:10 EST
Abstract: Against a background of growing interest in reproducibility in NLP and ML, and as part of an ongoing research programme designed to develop theory and practice of reproducibility assessment in NLP, we organised the second shared task on reproducibility of evaluations in NLG, ReproGen 2022. This paper describes the shared task, summarises results from the reproduction studies submitted, and provides further comparative analysis of the results. Out of six initial team registrations, we received submissions from five teams. Meta-analysis of the five reproduction studies revealed varying degrees of reproducibility, and allowed further tentative conclusions about what types of evaluation tend to have better reproducibility.