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

Role-playing Dialogue Evaluation on FURINA-Bench Chinese

71.39Context Reliance

Qwen3-32B

Updated 1mo ago

Evaluation Results

Method	Links
Qwen3-32B 2025.10		71.39	68.42	54.9	67.84	82.71	69.05
Deepseek-R1 2025.10		67.91	70.51	80.19	65.77	82.53	73.38
Qwen3-235B 2025.10		67.07	70.55	57.68	70	81.39	69.34
Qwen3-32B 2025.10		64.14	57.43	55.11	60.49	78.34	63.1
Qwen3-235B 2025.10		59.38	54.3	56.87	56.38	72.92	59.97
Qwen3-8B 2025.10		53.23	49.06	41.21	49.17	69.12	52.36
o3 2025.10		47.46	39.06	54.6	47.86	56.77	50.15
Qwen3-8B 2025.10		45.07	40.32	43.24	39.81	59.69	45.63
Claude-4-Sonnet 2025.10		40.17	37.62	69.75	36.57	48.56	46.33
Claude-4-Sonnet 2025.10		37.77	32.48	65.24	35.48	49.1	44.02
Deepseek-V3 2025.10		36.31	44.25	21.49	31.84	49.95	36.77
GPT-4o 2025.10		28.49	19.78	21.68	23.67	28.94	24.51