Dataset Viewer
Auto-converted to Parquet Duplicate
paper_id
stringlengths
10
10
messages
listlengths
3
3
pred_score
float64
1
9
avg_score
float64
1
10
scores
listlengths
2
6
mae
float64
0
5.8
is_success
bool
2 classes
04RGjODVj3
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
3.8
3
[ 1, 5, 3, 3 ]
0.8
true
04RGjODVj3
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
4
3
[ 1, 5, 3, 3 ]
1
true
04qx93Viwj
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
6.5
7.33
[ 8, 8, 6 ]
0.833333
true
04qx93Viwj
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
6.5
7.33
[ 8, 8, 6 ]
0.833333
true
07yvxWDSla
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
6
8
[ 8, 8, 8, 8 ]
2
false
0CvJYiOo2b
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
5.5
4.5
[ 5, 5, 3, 5 ]
1
true
0CvJYiOo2b
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
4.5
4.5
[ 5, 5, 3, 5 ]
0
true
0F1rIKppTf
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
5.8
5.75
[ 6, 5, 6, 6 ]
0.05
true
0F1rIKppTf
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
6
5.75
[ 6, 5, 6, 6 ]
0.25
true
0JwxMqKGxa
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
1.5
3.17
[ 6, 5, 1, 3, 3, 1 ]
1.666667
true
0JwxMqKGxa
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
1.5
3.17
[ 6, 5, 1, 3, 3, 1 ]
1.666667
true
0Q1mBvUgmt
[ { "role": "system", "content": "You are an experienced ML reviewer. You are given the text of a peer review of a paper. Based only on the review's content, judge how strong the paper is and predict the average score the human reviewers gave it, on the ICLR 1-10 scale (1 strong reject, 3 reject, 4 borderline...
5.5
3
[ 3, 3, 3 ]
2.5
false
End of preview. Expand in Data Studio
README.md exists but content is empty.
Downloads last month
38