Update README.md
Browse filesupdated model results
README.md
CHANGED
|
@@ -33,7 +33,7 @@ and found at [implicitmotives.com](https://implicitmotives.com). The classifier
|
|
| 33 |
presence of implicit motive imagery in sentences, namely the three felt needs for Power, Achievement,
|
| 34 |
and Affiliation.
|
| 35 |
|
| 36 |
-
The current classifier is finetuned from ELECTRA-base and achieves > 0.
|
| 37 |
Winter (1994) training data (see the [OSF repo](https://osf.io/aurwb/) for the benchmark dataset).
|
| 38 |
Development of this classifier is ongoing, and the current version has been trained on a larger and
|
| 39 |
more diverse dataset, which means it generalizes better to unseen data.
|
|
@@ -46,16 +46,16 @@ like to use this model commercially, get in touch with us for access to our most
|
|
| 46 |
Predictions on Winter manual dataset
|
| 47 |
-----
|
| 48 |
Intra-class Correlation Coefficient:
|
| 49 |
-
| Pow (Label_0): | 0.
|
| 50 |
-
| Ach (Label_1): | 0.
|
| 51 |
-
| Aff (Label_2): | 0.
|
| 52 |
-
| mean: | 0.
|
| 53 |
|
| 54 |
Pearson correlations:
|
| 55 |
-
| Pow (Label_0): 0.
|
| 56 |
-
| Ach (Label_1): 0.
|
| 57 |
-
| Aff (Label_2): 0.
|
| 58 |
-
| mean: 0.
|
| 59 |
|
| 60 |
```
|
| 61 |
|
|
|
|
| 33 |
presence of implicit motive imagery in sentences, namely the three felt needs for Power, Achievement,
|
| 34 |
and Affiliation.
|
| 35 |
|
| 36 |
+
The current classifier is finetuned from ELECTRA-base and achieves > 0.90 ICC on the
|
| 37 |
Winter (1994) training data (see the [OSF repo](https://osf.io/aurwb/) for the benchmark dataset).
|
| 38 |
Development of this classifier is ongoing, and the current version has been trained on a larger and
|
| 39 |
more diverse dataset, which means it generalizes better to unseen data.
|
|
|
|
| 46 |
Predictions on Winter manual dataset
|
| 47 |
-----
|
| 48 |
Intra-class Correlation Coefficient:
|
| 49 |
+
| Pow (Label_0): | 0.90469 |
|
| 50 |
+
| Ach (Label_1): | 0.93134 |
|
| 51 |
+
| Aff (Label_2): | 0.88893 |
|
| 52 |
+
| mean: | 0.90815 |
|
| 53 |
|
| 54 |
Pearson correlations:
|
| 55 |
+
| Pow (Label_0): 0.81604 |
|
| 56 |
+
| Ach (Label_1): 0.85726 |
|
| 57 |
+
| Aff (Label_2): 0.77257 |
|
| 58 |
+
| mean: 0.81455 |
|
| 59 |
|
| 60 |
```
|
| 61 |
|