List of ICLR 2018 submissions ranked by number of replies

iclr18-ranking-by-replies.sh

# requirements: curl, jq, python, pandas

# brew install jq
# pip install pandas

# prints top 50 papers

curl -sS "https://openreview.net/notes?invitation=ICLR.cc%2F2018%2FConference%2F-%2FBlind_Submission&offset=0&limit=5000" | jq -r '.notes[] | .content.title + "\t" + (.replyCount|tostring)' > subset.tsv && python -c 'from pprint import pprint; import pandas as pd; d=pd.read_csv("subset.tsv", header=None, delimiter="\t"); d.columns=["title", "replies"]; pprint(d.sort_values(by="replies", ascending=False).head(50).title.tolist());' && rm -f subset.tsv

# number of papers in top 50 on adversarial networks

curl -sS "https://openreview.net/notes?invitation=ICLR.cc%2F2018%2FConference%2F-%2FBlind_Submission&offset=0&limit=5000" | jq -r '.notes[] | .content.title + "\t" + (.replyCount|tostring)' > subset.tsv && python -c 'from pprint import pprint; import pandas as pd; d=pd.read_csv("subset.tsv", header=None, delimiter="\t"); d.columns=["title", "replies"]; pprint(d.sort_values(by="replies", ascending=False).head(50).title.tolist());' | grep -ic 'adversarial\|GAN' && rm -f subset.tsv