SamSum
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
SAMSum 1.0 (test)
51R1
11
SAMSum
50.78ROUGE-1
10
Samsum
99.99AUROC
10
SAMSum
29.88ROUGE-2
10
Samsum
54.1BRT Score
9
Samsum
4.02PPL
9
SamSum
46.4PRR
8
SAMSum
54.27ROUGE-1
8
SamSum Out-of-domain
45.2PRR (ROUGE-L)
8
Samsum
1AUROC
8
SamSum
47.7Pearson Correlation Coefficient
8
SamSum
38.2Kendall's Tau
8
SAMSum (val)
53.8ROUGE-1
8
SAMSum
44.59ROUGE-1
7
SAMSum
0.523MiniCheck Score
7
SAMSum Multiple Client (test)
49.99ROUGE-1 (Client 1)
6
SAMSum Single Client
50.59ROUGE-1
6
SamSum 4.8:1
6.14Cohen's d
6
SAMSum (test)
85.5R@15
6
SAMSum 200 samples (test)
4.94Faithfulness
6
SAMSum 30 samples (test)
4.52Faithfulness
6
Samsum
94Monitor Accuracy
4
SAMSum
75.4AUC
4
Samsum 4k~8k
0.428BLEURT Score
4
Samsum 2k~4k
0.433BLEURT
4