Scientific Document Retrieval on CSFCube (test) benchmark leaderboard · SOTA2 Research

Scientific Document Retrieval on CSFCube (test)

20.85M@10CCQGen

Evaluation Results

Method	Links
CCQGen2026.01		20.85	—	—	25.49	58.86	73.55	41.05	41.76
CCQGen2026.01		20.4	—	—	25.34	58.44	73.33	40.31	41.5
Pair-wise generation2026.01		19.66	—	—	24.23	57.22	69.72	38.7	39.99
Promptgator2026.01		17.9	—	—	22.45	57.15	69.62	37.66	38.86
Promptgator_diverse2026.01		17.21	—	—	21.57	56.87	69.72	36.72	38.01
GenQ2026.01		16.99	—	—	20.83	55.41	68.36	36.58	36.59
CCQGen2026.01		16.56	—	—	22.28	63.62	75.26	36.7	40.63
CCExpand2026.01		16.39	36.41	37.4	21.42	55.71	69.48	—	—
Doc exp. (Ours)2026.01		16.15	34.98	37.59	20.74	58.09	70.46	—	—
Retriever2026.01		16.15	35.03	35.79	20.43	53.41	68.59	—	—
no Fine-Tune2026.01		16.15	—	—	20.43	53.41	68.59	35.03	35.79
CCQGen2026.01		16.14	—	—	21.94	63.33	74.67	36.05	39.91
Promptgator_diverse2026.01		16.06	—	—	20.29	59.5	71.32	35.39	37.71
CCExpand2026.01		15.79	34.12	37.08	20.01	60.08	71.41	—	—
CCExpand--2026.01		15.53	34.22	36.39	20.86	53.81	69.99	—	—
Promptgator2026.01		15.38	—	—	19.74	59.28	72.98	34.41	36.7
CCExpand--2026.01		15.32	32.68	35.13	19.04	58.29	71.05	—	—
Doc exp. (Ours)2026.01		15.3	34.27	35.67	19.97	53.46	68.54	—	—
Retriever2026.01		15.25	33.13	36.04	19.37	57.83	71.36	—	—
no Fine-Tune2026.01		15.25	—	—	19.37	57.83	71.36	33.13	36.04
Pair-wise generation2026.01		15.22	—	—	19.71	59.61	72.25	34.18	36.86
Doc exp. (BestQGen)2026.01		15.11	33.41	36.31	19.48	56.89	68.68	—	—
GenQ2026.01		14.76	—	—	18.41	55.71	68.43	34.01	34.95
GRF2026.01		13.89	30.75	32.06	17.87	53.25	64.95	—	—
GRF2026.01		13.88	30.47	33.21	17.8	55.76	66.97	—	—
Doc exp. (BestQGen)2026.01		11.76	29.58	30.39	15.07	53.42	65.09	—	—