Hurdles to Progress in Long-form Question Answering
Open paper
Benchmark
Model
Rank
Results
open-domain-question-answering-on-kilt-eli5
arxiv.org/abs/2103.06332
#1
KILT-RL: 2.36
R-Prec: 10.67
Recall@5: 24.56
ROUGE-L: 23.19
…
question-answering-on-kilt-eli5
c-REALM
#2
Rouge-L: 23.4
F1: 23.1
Rank counts only results with a code link.