Information Extraction on 1,000-note gold-standard set 1.0 (test) benchmark leaderboard · SOTA2 Research

Information Extraction on 1,000-note gold-standard set 1.0 (test)

94.8Micro PrecisionQwen2.5-14B-Instruct

Evaluation Results

Method	Links
Qwen2.5-14B-Instruct2026.05		94.8	79.4	86.4	94.8	76.9	83.7
Llama-3.1-8B-Instruct2026.05		94.8	70.1	80.6	94.9	69.6	79.7
Qwen2.5-7B-Instruct2026.05		92.5	70.5	80	90.8	68.3	76.9
gemma-2-9b-it2026.05		91	1	2	62.6	2.1	3.5
OLMo-2-1124-7B-DPO2026.05		79.9	64.6	71.4	78.8	58.4	64.6
Qwen2.5-14B-Instruct2026.05		79.8	76.2	78	79.1	74.8	75.5
Qwen2.5-7B-Instruct2026.05		73.6	71	72.3	74.2	71.9	71.5
Llama-3.1-8B-Instruct2026.05		69.4	83.8	75.9	69.3	85.3	73.8
granite-3.1-8b-instruct2026.05		64.6	63.9	64.3	69	66.3	66.1
OLMo-2-1124-7B-DPO2026.05		63.1	70.6	66.6	57.4	58.3	55.4
OLMo-2-1124-7B-DPO2026.05		61.7	32.9	42.9	54.6	28.6	35.1
Qwen2.5-14B-Instruct2026.05		61.6	89.5	73	62.3	88.1	70.9
gemma-2-9b-it2026.05		57.5	63.7	60.4	65	64.9	62.2
Qwen2.5-7B-Instruct2026.05		57.1	82.3	67.4	61	83.3	66.9
OLMo-2-1124-13B-DPO2026.05		56.5	72.5	63.5	65.2	69.6	64.6
Llama-3.1-8B-Instruct2026.05		52.3	76.1	62	56.3	78.3	61.5
gemma-2-9b-it2026.05		51	80.3	62.4	58.8	75.9	63
Llama-3.2-3B-Instruct2026.05		44.9	59.2	51.1	42.2	51.3	42.4
OLMo-2-1124-13B-DPO2026.05		43.3	86.3	57.7	46.3	84.2	57.2
Mistral-Small-Instruct-24092026.05		40.5	62.8	49.2	42.4	71.7	47
granite-3.1-8b-instruct2026.05		38.2	89.2	53.5	43.9	89.2	54.7
Mistral-Small-Instruct-24092026.05		31.8	43	36.5	44.8	45.8	37
Mistral-7B-Instruct-v0.32026.05		29.6	37.8	33.2	36.1	29.6	26.7
Llama-3.2-3B-Instruct2026.05		25.1	46.6	32.6	36.1	46.7	35.3
Mistral-7B-Instruct-v0.32026.05		15.8	86.8	26.7	21.6	84.6	30.7