MIT Press

Table 12:

Cross-model micro-averaged accuracy (%). The first row is the model to test, and the second row is the model on which prompt weights are learned.

Test	BERT-base		BERT-large
Train	base	large	large	base
Mine	38.9	38.7	43.7	42.2
Mine+Man	39.6	40.1	43.9	42.2
Mine+Para	36.2	35.6	40.1	39.0
Man+Para	37.3	35.6	38.8	37.5

Test	BERT-base		BERT-large
Train	base	large	large	base
Mine	38.9	38.7	43.7	42.2
Mine+Man	39.6	40.1	43.9	42.2
Mine+Para	36.2	35.6	40.1	39.0
Man+Para	37.3	35.6	38.8	37.5