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

Multi-hop Question Answering on Multi-Hop QA (Aggregated)

89.342Wiki Accuracy

Gemini-2.5-Flash-Thinking

Updated 1mo ago

Evaluation Results

Method	Links
Gemini-2.5-Flash-Thinking 2026.01		89.34	81	60	76.78
GPT5-Nano 2026.01		89	82	61.5	77.5
GPT-OSS-120B 2026.01		89	82	66	79
GPT-OSS-20B 2026.01		88.5	79	49.25	72.25
LONGPAS 2026.01		86.5	75.38	55	72.29
LONGPAS 2026.01		86.5	77	66	76.5
RLVR 2026.01		86	76	57	73
LONGPAS 2026.01		85.62	78.62	54.25	72.83
Qwen3-4B-Thinking 2026.01		85.5	75.12	50.25	70.29
LONGPAS 2026.01		84.5	77.5	55.5	72.5
Qwen3-30B-A3B-Thinking 2026.01		84.31	76.94	63.44	74.9
RLVR 2026.01		84	77	64	75
Qwen3-30B-A3B-Instruct 2026.01		83.88	77.62	51.5	71
RLVR 2026.01		83.5	71.5	53.5	69.5
RLVR 2026.01		82.62	75.88	49.25	69.25
LONGPAS 2026.01		80.88	80	51.62	70.83
LONGPAS 2026.01		79.38	76.38	52	69.25
RLVR 2026.01		68.25	67.75	40.25	58.75
RLVR 2026.01		68	69.75	46.62	61.46
Qwen3-4B-Instruct 2026.01		63.25	67.62	28.75	53.21
AtomicRAG 2026.02		56.8	70.5	50.9	59.4
Qwen2.5-7B-Instruct 2026.01		54.75	69.75	32.38	52.29
StructRAG 2026.02		52.3	49.2	24.9	42.1
RAPTOR 2026.02		50.1	66.4	43.6	53.4
Fast-GraphRAG 2026.02		47.8	62.3	42.1	50.7
HippoRAG2 2026.02		47.5	67.5	41.5	52.2
LightRAG 2026.02		44.2	57	23.5	41.6
LLaMA3.1-8B-Instruct 2026.01		43.62	61.75	21	42.12
GFM-RAG 2026.02		43.6	68.6	39.9	50.7
Lazy-GraphRAG 2026.02		40.4	54.6	39.1	44.7
MS-GraphRAG 2026.02		39.5	37.6	35.4	37.5
KGP 2026.02		38.6	52.3	37.2	42.7
MS-GraphRAG 2026.02		37.4	52.2	37.8	42.5
RAG 2026.02		32.1	54.5	33.3	40
RAG 2026.02		31.8	54	32.2	39.3
HippoRAG 2026.02		29.8	38.1	25.7	31.2
KET-RAG 2026.02		24.5	45.6	22.6	30.9