Table 9.

p-values from Mann–Whitney U test comparing the correctness of the results obtained for temperature parameter equal to 0 and 1.

	S22	A22	S23
Polish
GPT-3.5	0.369 ns	0.161 ns	0.108 ns
GPT-4	1.000 ns	0.462 ns	0.710 ns
English
GPT-3.5	0.763 ns	0.839 ns	0.052 ns
GPT-4	0.566 ns	0.902 ns	0.389 ns