Table 3.
Comparison of Fleiss’ κ scores with scores from SNLI quality control sentence pairs.
Fleiss’ κ | 4GS | 5GS | Bowman et al. 2015 |
---|---|---|---|
Contradiction | 0.37 | 0.59 | 0.77 |
Entailment | 0.48 | 0.63 | 0.72 |
Neutral | 0.41 | 0.54 | 0.6 |
Overall | 0.43 | 0.6 | 0.7 |
Comparison of Fleiss’ κ scores with scores from SNLI quality control sentence pairs.
Fleiss’ κ | 4GS | 5GS | Bowman et al. 2015 |
---|---|---|---|
Contradiction | 0.37 | 0.59 | 0.77 |
Entailment | 0.48 | 0.63 | 0.72 |
Neutral | 0.41 | 0.54 | 0.6 |
Overall | 0.43 | 0.6 | 0.7 |