Factuality Checking on LLM-AggreFact (test) benchmark leaderboard · SOTA2 Research

Factuality Checking on LLM-AggreFact (test)

72.5CNN ScoreQwen3-30B-A3B-Instruct

Evaluation Results

Method	Links
Qwen3-30B-A3B-Instruct2026.05		72.5	74.4	70.6	81.8	77	88.4	72.7	79.4	60.6	79	82.6	76.3	4
Qwen3-30B-A3B-Thinking2026.05		71.7	72.9	73.6	78.2	69.8	85.7	73.2	72.7	58.1	81.2	82.1	74.5	8
Qwen3-4B-Instruct2026.05		70.8	70	73.7	79.3	75.4	88.3	72.9	75.3	59.1	82.4	84	75.6	5.3
FactCG2026.05		70.1	73.9	72.3	74.3	74.2	88.4	78.5	72.1	59.1	86.7	82.3	75.6	5.8
MiniCheck-FT52026.05		69.9	74.3	73.6	77.3	72.2	86.2	74.6	74.7	59	85.2	78	75	6.8
Qwen3-0.6B-Instruct + SFT + DPO2026.05		69.8	72.2	68.7	75.8	75.4	86.1	71.3	75.2	58.7	77.9	77.1	73.6	9.5
Llama-3.3-70B-Instruct2026.05		68.7	74.7	69.5	78.4	76.6	85.5	67.4	78.5	58.3	79.8	82.6	74.5	7.3
GPT-4o-2024-05-132026.05		68.1	76.8	71.4	79.8	78.5	86.5	69	77.5	59.6	83.6	84.3	75.9	4.8
Qwen3-4B-Thinking2026.05		68	69.8	69.8	75.3	73.1	84.6	71.5	74	57.3	77.8	81	73.1	11
MiniCheck-BeSpoke-7B2026.05		65.5	77.8	76	78.3	83	88	75.3	77.7	59.2	86.7	84	77.4	3.3
Qwen3-0.6B-Instruct + SFT2026.05		65.1	71.3	66.2	72.4	73	86.4	66.2	73.6	58.7	75.7	75.9	71.3	12
MiniCheck-DeBERTa2026.05		64.2	71	69.3	72.7	69.4	87.3	75.6	73	58.9	83.9	78.8	73.1	10
MiniCheck-RoBERTa2026.05		63.7	70.8	71.9	75.9	67.6	88.8	77.4	73.3	57.4	84.4	77.2	73.5	9.1
ChatGPT-ZS2026.05		63.2	72.4	66.8	73.4	68.5	84.7	65.2	70.8	57.2	73.8	75.6	70.1	13.8
TrueTeacher2026.05		60.4	74.2	70.9	73.6	64.2	91.1	64.4	76.8	59.5	90.2	80.9	73.3	8.4
Qwen3-0.6B-Instruct2026.05		51.5	51.6	55.2	55.4	52.1	70.4	54	61.3	52.3	58.2	57.6	56.3	16