Table A2. Single-dataset classifiers, filtered.

Instances where training and test data belong to the same language model are highlighted (bold).

Training data	Test data
	s		xl		s-k		xl-k		GPT3		Grover
	Acc.	AUC	Acc.	AUC	Acc.	AUC	Acc.	AUC	Acc.	AUC	Acc.	AUC
s	0.930	0.982	0.769	0.884	0.473	0.307	0.459	0.273	0.320	0.2139	0.431	0.43
xl	0.849	0.971	0.802	0.883	0.446	0.329	0.426	0.303	0.387	0.328	0.494	0.477
s-k	0.321	0.172	0.443	0.292	0.947	0.985	0.801	0.939	0.609	0.812	0.505	0.667
xl-k	0.216	0.099	0.360	0.242	0.910	0.974	0.861	0.933	0.637	0.660	0.514	0.721
GPT3	0.417	0.131	0.432	0.254	0.806	0.884	0.734	0.820	0.754	0.834	0.614	0.668
Grover	0.334	0.286	0.423	0.395	0.764	0.842	0.711	0.762	0.731	0.747	0.676	0.769