Share your thoughts, 1 month free Claude Pro on usSee more

Preference Evaluation on VCD

94.15Multi-choice One Preference p+

SFT

Updated 4d ago

Evaluation Results

Method
SFT 2026.04	94.15	64.37	79.26	62.1	34.74	48.42	6.61	8.43
PFT 2026.04	91.71	86.04	88.88	64.35	60.31	62.33	7.26	8.61
CPO 2026.04	91.2	66.8	79	58.4	36.1	47.25	6.55	8.38
SFT 2026.04	90.1	83.13	86.62	60.05	51.4	55.72	6.76	8.48
CPO 2026.04	89.5	84.1	86.8	60.25	52.8	56.53	6.85	8.45
DPO 2026.04	89.29	79.53	84.41	57.71	44.66	51.18	7.01	8.26
DPO 2026.04	88.48	85.54	87.01	59.91	53.69	56.8	7.03	8.43
PFT 2026.04	88.29	81.84	85.07	47.1	42.43	44.76	7.53	8.57
SFT 2026.04	85.29	35.32	60.3	32.33	11.36	21.84	7.15	7.81
SFT 2026.04	85.06	74.2	79.63	24.82	15.51	20.16	6.44	8.41
CPO 2026.04	84.1	75.5	79.8	28.5	18.2	23.35	6.5	8.42
CPO 2026.04	83.45	38.12	60.79	30.15	10.88	20.52	6.95	7.72
CAA 2026.04	81.11	71.71	76.41	54.58	49.21	51.89	6.65	8.29
DPO 2026.04	77.75	73.94	75.85	41.13	39.62	35.17	6.47	8.44
PFT 2026.04	77.54	72.38	74.96	53.57	52.12	52.84	7.18	8.64
DPO 2026.04	75.44	44.79	60.12	26.77	4.63	15.7	6.82	7.65
CAA 2026.04	72.73	81.58	77.16	50.13	54.24	52.18	6.5	8.36
DPO 2026.04	72.23	74.4	73.32	52.24	49.29	50.76	6.87	8.61
DPO 2026.04	72.23	84.79	78.51	45.56	53.12	49.34	6.74	8.15
SFT 2026.04	70.38	58.7	64.54	52.38	38.41	45.4	6.79	8.03
CPO 2026.04	68.9	60.15	64.53	50.8	39.5	45.15	6.65	8.12
CPO 2026.04	68.3	85.4	77.35	54.2	50.15	52.18	6.18	8.15
SFT 2026.04	66.64	88.25	77.45	53.07	52.1	52.58	6.06	8.02
CAA 2026.04	66.34	59.49	62.92	47.56	41.13	44.35	6.32	8.3
DPO 2026.04	65.07	69.99	67.53	51.01	48.99	50	6.84	8.31
CAA 2026.04	60.97	65.55	63.26	20.66	24.35	22.51	5.82	7.83
CPO 2026.04	60.45	68.1	64.28	51.1	42.35	46.73	6.45	8.5
Base 2026.04	57.92	74.08	66	41.47	43.8	42.64	6.08	8.59
CAA 2026.04	52.2	70.75	61.48	42.65	46.55	44.6	6	8.44
SFT 2026.04	52.11	67.52	59.82	51.91	40.46	46.18	6.3	8.43
CAA 2026.04	50.75	76.06	63.41	15.68	33.44	24.56	5.63	8.05
Base 2026.04	48.3	63.51	55.91	11.68	11.59	11.64	6.54	8.4
Base 2026.04	48.27	55.48	51.88	48.35	48.58	48.46	6.04	8.27
CPO 2026.04	36.5	78.2	57.35	44.12	38.9	41.51	5.8	8.48
DPO 2026.04	35.77	77.29	56.53	25.73	45.63	35.68	5.16	8.66
CPO 2026.04	35.12	83.9	59.51	22.8	30.05	26.43	6.2	8.35
SFT 2026.04	34.39	80.42	57.41	46.34	36.21	41.27	5.67	8.43
SFT 2026.04	33.68	85.61	59.65	24.26	31.78	28.02	6.15	8.42
DPO 2026.04	27.5	79.66	53.58	20.16	26.16	23.16	6.44	8.61