Sentiment classification on RoIt-XMASA benchmark leaderboard · SOTA2 Research

Sentiment classification on RoIt-XMASA

62.45Accuracy (Books)DeepSeek-R1-8B

Evaluation Results

Method	Links
DeepSeek-R1-8B2026.04		62.45	57.81	63.34	59.12	61.23	55.67	65.12	61.23	61.89	57.04	62.01	56.8
Llama-3.1-8B2026.04		61.89	56.78	66.23	61.85	63.12	58.15	66.34	60.89	61.15	56.63	63.66	58.43
DeepSeek-R1-8B2026.04		59.78	54.89	63.67	58.67	61.29	55.49	64.45	58.92	58.71	53.78	61.58	56.35
Qwen3-8B2026.04		59.67	54.23	61.45	56.12	63.78	59.34	63.45	57.92	59.82	55.11	60.11	54.97
Llama-3.1-8B2026.04		58.23	49.12	62.45	53.68	60.08	50.08	63.12	54.23	57.34	47.69	60.25	50.96
Qwen3-8B2026.04		56.45	48.23	60.34	52.78	58.05	49.26	61.12	52.89	55.44	47.29	58.28	50.09
Mistral-7B-v0.32026.04		53.89	47.23	57.45	51.34	54.56	48.4	57.89	51.12	53.71	46.86	55.3	48.99
Mistral-7B-v0.32026.04		52.34	46.12	56.23	50.12	53.97	47.85	56.78	50.23	51.58	45.83	54.18	48.03