k | \(PPL_{err}\) | \(PPL_{corr}\) | \(PPL_{total}\) |
---|
15 | 1073.6 | 943.5 | 952.9 |
37 | 1072.8 | 944.5 | 946.8 |
81 | 1072.8 | 973.4 | 956.3 |
- \(PPL_{err}\), \(PPL_{corr}\), and \(PPL_{total}\) denote the perplexity scores on erroneous and error-free reads, and the entire dataset (i.e., erroneous and error-free sequences)