Multi-hop Question Answering

Benchmarks

Dataset Name	SOTA Method	Metric
2WikiMultiHopQA	SynPlanResearch-R1	EM82.1	559	2mo ago
HotpotQA (test)		F180.79	334	18d ago
HotpotQA	SAFE	F1 Score77.4	294	3mo ago
2WikiMultiHopQA (test)	CJ	EM73.9	247	25d ago
2Wiki	G-reasoner	Exact Match74.9	215	2mo ago
Musique	INSES + Router	EM46	209	26d ago
2WikiMQA	CIRAG	F1 Score76.4	175	22d ago
HotpotQA	Tree-GRPO	Exact Match (EM)50.2	167	1mo ago
MuSiQue (test)	SEARCH-R	F155.68	151	18d ago
Bamboogle	EvalAct	Exact Match56	128	3mo ago
LoCoMo	Membox	F148.35	125	1mo ago
Bamboogle (test)	ARBOR	EM67.5	110	1mo ago
2WikiMHQA	CRAFT7B	F1 Score85.56	85	1mo ago
HotpotQA	CIRAG	F174.9	79	4mo ago
Multi-hop RAG	IndexLM-4B	F186.07	77	3mo ago
HotpotQA	INSES + Router	LLM Judge Score80	72	1mo ago
2WikiMultiHopQA Out-Of-Distribution (OOD)	QwenLong-L1-32B	Accuracy74.2	72	4mo ago
HotpotQA (dev)	Dense-aware HKVM controller	Answer F185.81	72	25d ago
2WikiQA (test)	MultiCube-RAG	F174.3	71	2mo ago
HotpotQA	SD-Search-Instruct	Exact Match (EM)47.1	66	2mo ago
Multi-Hop QA (HotpotQA, 2Wiki, Musique, Bamboogle) (test)	JADE	HotpotQA Score57.02	65	1mo ago
Multi-Hop QA (HotpotQA, 2Wiki, Musique, Bamboogle)	Qwen2.5-32B-Instruct + Proposed Method	HotpotQA Score64.6	64	1mo ago
HotpotQA fullwiki setting (test)	IDRQA	Answer F175.9	64	4mo ago
Bamboogle	TC+FM*	Accuracy75.2	62	4mo ago
MuSiQue	EvalAct	Exact Match (EM)25.3	58	3mo ago

Showing 25 of 512 rows

...