Distractor Generation on MCQL benchmark leaderboard · SOTA2 Research

Distractor Generation on MCQL

36.17P@1GPT-3(few-shot-COT-k-NN)

Evaluation Results

Method	Links
GPT-3(few-shot-COT-k-NN)2026.04		36.17	12.06	23.61	39.94	42.9	0	0
GPT-3(few-shot-k-NN)2026.04		35.33	11.78	24.44	39.61	42.81	0	0.33
GPT-3(few-shot-COT-random)2026.04		32	10.67	19.44	35.58	38.51	0	0.17
GPT-3(few-shot-random)2026.04		30.17	10.06	19.06	34.36	37.98	0	0.5
TinyLlama2026.04		28.83	9.61	20.56	33.33	37.03	2.67	1.17
GPT-3(zero-shot)2026.04		24.17	8.06	13.78	27.03	29.37	0	0.17
T5(contrast)2026.04		23	7.67	13.67	26.42	29.13	16.67	3.67
T5-CSG-DS(beam)2026.04		20.33	6.78	14.5	25.42	29.65	37.17	1.83
T5(multi-task)2026.04		19.33	6.44	11.5	22.14	24.45	21.5	4.17
T5(base)2026.04		17.83	5.94	10.78	20.07	21.68	28.83	4.83
T5-CSG-DS(clustering)2026.04		12.17	4.06	5.56	15.69	19.04	20.5	0
GPT-3-CSG-DS(clustering)2026.04		11.17	3.72	4.28	13.92	16.55	0.17	0