Summ
Benchmarks
TaskDatasetSOTA resultTrendResults
Summ
0.517Mean PRR
109
Summ
81.5ROC-AUC
64
Summ
100Detection Rate (TPR/FPR)
8
Summ
5.69BLEU-4
5
Summ Qwen2-7B-Instruct v1 (test)
1.5Acceptance Length (τ)
4
Summ
0.962RL Score
3
Summ.
182.7Throughput (tokens/s)
3