The performance of an RNN classifier using h alone as input, in comparison to the random baseline. Missing cells denote cases where we were unable to converge training.
| Model . | SST-2 . | AGNews . | IMDB . | Ev.Inf. . | 20News . | Elec . |
|---|---|---|---|---|---|---|
| Random baseline | 50.0 | 25.0 | 50.0 | 33.33 | 5.0 | 50.0 |
| Lei et al. (2016) | 59.7 | 41.4 | 69.11 | 33.45 | 60.75 | |
| Bastings et al. (2019) | 62.8 | 42.4 | 9.45 | |||
| FRESH | 52.22 | 35.35 | 54.23 | 38.88 | 11.11 | 58.38 |
| Model . | SST-2 . | AGNews . | IMDB . | Ev.Inf. . | 20News . | Elec . |
|---|---|---|---|---|---|---|
| Random baseline | 50.0 | 25.0 | 50.0 | 33.33 | 5.0 | 50.0 |
| Lei et al. (2016) | 59.7 | 41.4 | 69.11 | 33.45 | 60.75 | |
| Bastings et al. (2019) | 62.8 | 42.4 | 9.45 | |||
| FRESH | 52.22 | 35.35 | 54.23 | 38.88 | 11.11 | 58.38 |