List your company

Multi-hop Question Answering on Musique

46EMINSES + Router
13.86422.20730.5538.893Oct 7, 2022May 16, 2023Dec 24, 2023Aug 2, 2024Mar 12, 2025Oct 20, 2025May 30, 2026

Evaluation results

MethodLinks
INSES + RouterCategory=Ours2026.0346———————47————
SiReRAGCategory=Graph-based2026.0344———————43————
ARISEBackbone=Qwen2.5-14B-Instruct, Method Strategy=Search-based2025.0440.565.87———————————
LightRAGCategory=Graph-based2026.0338———————42————
SynPlanResearch-R1Model Backbone=Qwen3-8B2026.0335.7————————————
Auto-RAGBackbone=Qwen2.5-14B-Instruct, Method Strategy=Prompt-based2025.0435.559.05———————————
SynPlanResearch-R1Model Backbone=Qwen3-4B2026.0333.3————————————
Naïve RAGCategory=Text-based, Top-k=102026.0333———————37————
SimpleDeepSearcherModel Backbone=Qwen3-8B2026.0332.3————————————
RAPTORCategory=Graph-based2026.0332———————35————
Naïve RAGCategory=Text-based, Top-k=52026.0331———————29————
CARLModel Scale=7B-Inst.2026.0530.4—————————2,511——
SimpleDeepSearcherModel Backbone=Qwen3-4B2026.0329.8————————————
Auto-RAGBackbone=Qwen2.5-7B-Instruct, Method Strategy=Prompt-based2025.0429.557.19———————————
RATTBackbone=Llama3.1-8B-Instruct, Method Strategy=Search-based2025.0429.563.66———————————
ARISEBackbone=Qwen2.5-7B-Instruct, Method Strategy=Search-based2025.042962.26———————————
Search-R1Model Backbone=Qwen3-8B2026.0328.8————————————
Search-R1Model Backbone=Qwen3-4B2026.0326.5————————————
GraphRAGCategory=Graph-based, Top-k=102026.0326———————36————
Tree-GRPOBackbone=Qwen2.5-14b2025.0925.9————————————
Self-AskBackbone=Qwen2.5-14B-Instruct, Method Strategy=Prompt-based2025.042558.59———————————
IRCoTCategory=Text-based2026.0325———————42————
ARISEBackbone=Llama3.1-8B-Instruct, Method Strategy=Search-based2025.0424.566.38———————————
RATTBackbone=Qwen2.5-14B-Instruct, Method Strategy=Search-based2025.042463.76———————————
GLM-4Category=LLM only, Protocol=Few-shot CoT2026.0324———————27————
GSPOBackbone=Qwen2.5-14b2025.0923.8————————————
Critic-Embedtop k=5, Reasoner=Search-R12026.0523.4632.77———————————
Rejection SamplingModel Backbone=Qwen3-8B2026.0323.4————————————
Stella-400Mtop k=5, Reasoner=Search-R12026.0523.2132.49———————————
GRPOBackbone=Qwen2.5-14b2025.0923.2————————————
GraphRAGCategory=Graph-based, Top-k=52026.0323———————24————
StepSearchBackbone=Qwen2.5-7b-Instruct2026.0122.6————————————
Agentic-RSearch Agent=R1-Searcher, Search Agent Domain=out-of-domain2026.0122.54————————————
Critic-Embedtop k=3, Reasoner=Search-R12026.0522.4231.85———————————
Agentic-Rtop k=5, Reasoner=Search-R12026.0522.3831.48———————————
Dep-SearchBackbone=Qwen2.5-7B-Instruct2026.0122.2————————————
MemPO2026.0222.133.48————0.3915—————
Rejection SamplingModel Backbone=Qwen3-4B2026.0322.1————————————
Search-R1 PPOModel Scale=7B-Inst.2026.0522.1—————————2,682——
Query2DocBackbone=Qwen2.5-14B-Instruct, Method Strategy=Prompt-based2025.042255.79———————————
Verify-and-EditBackbone=Qwen2.5-14B-Instruct, Method Strategy=Prompt-based2025.042255.3———————————
Self-AskBackbone=Qwen2.5-7B-Instruct, Method Strategy=Prompt-based2025.042257.48———————————
CoT-SCBackbone=Llama3.1-8B-Instruct, Method Strategy=Search-based2025.042259.43———————————
Search-o1Model Backbone=Qwen3-8B2026.0321.6————————————
Verify-and-EditBackbone=Qwen2.5-7B-Instruct, Method Strategy=Prompt-based2025.0421.554.75———————————
Auto-RAGBackbone=Llama3.1-8B-Instruct, Method Strategy=Prompt-based2025.0421.552.61———————————
MT-PPOBackbone=Qwen2.5-7b-Instruct2026.0121.5————————————
GRPOmemory=none2026.0221.533.78————0.8633—————
E5Search Agent=R1-Searcher, Search Agent Domain=out-of-domain2026.0121.39————————————
Agentic-Rtop k=3, Reasoner=Search-R12026.0521.2730.13———————————
CARLModel Scale=3B-Inst.2026.0521.1—————————2,546——
HyDECategory=Text-based2026.0321———————31————
Stella-400Mtop k=3, Reasoner=Search-R12026.0520.8529.88———————————
Self-AskBackbone=Llama3.1-8B-Instruct, Method Strategy=Prompt-based2025.0420.552.14———————————
HierSearchBackbone=Qwen2.5-7B-Instruct2026.0120.4————————————
Search-o1Model Backbone=Qwen3-4B2026.0320.4————————————
Agentic-RSearch Agent=Our Search Agent, Search Agent Domain=in-domain2026.0120.27————————————
Tree-GRPOBackbone=Qwen2.5-7b2025.0920.2————————————
O2-SearcherBackbone=Qwen2.5-7B-Instruct2026.0120————————————
GSPOBackbone=Qwen2.5-7b2025.0919.2————————————
GRPOBackbone=Qwen2.5-7b2025.0919.1————————————
Critic-Embedtop k=1, Reasoner=Search-R12026.0519.0728.21———————————
Query2DocBackbone=Llama3.1-8B-Instruct, Method Strategy=Prompt-based2025.041949.91———————————
REPLUGSearch Agent=Our Search Agent, Search Agent Domain=in-domain2026.0118.9————————————
CoDA-BaseBackbone=Qwen2.5-3B, Retrieval Protocol=w/ Multi-Hop Retrieval2025.1218.7————————————
R1-SearcherBackbone=Qwen2.5-7B-Instruct2026.0118.6————————————
SCARLetSearch Agent=R1-Searcher, Search Agent Domain=out-of-domain2026.0118.53————————————
Search-R1 GRPOModel Scale=7B-Inst.2026.0518.5—————————2,579——
Stella-400Mtop k=1, Reasoner=Search-R12026.0518.4926.84———————————
ZeroSearchBackbone=Qwen2.5-7b-Instruct2026.0118.4————————————
LLM-EmbedderSearch Agent=R1-Searcher, Search Agent Domain=out-of-domain2026.0118.36————————————
BGESearch Agent=R1-Searcher, Search Agent Domain=out-of-domain2026.0118.36————————————
Agentic-Rtop k=1, Reasoner=Search-R12026.0518.3327———————————
REPLUGSearch Agent=R1-Searcher, Search Agent Domain=out-of-domain2026.0118.32————————————
TaSR-RAGBackbone=Qwen2.5-72B-Instruct2026.0318.324.4———————————
RATTBackbone=Qwen2.5-7B-Instruct, Method Strategy=Search-based2025.041856.58———————————
ReSearch2026.0217.829.63————0.5829—————
RAGModel Backbone=Qwen3-8B2026.0317.8————————————
Tree-GRPOBackbone=Qwen2.5-3b2025.0917.8————————————
E5Search Agent=Our Search Agent, Search Agent Domain=in-domain2026.0117.74————————————
Tree-GRPOBackbone=Llama3.2-3b2025.0917.6————————————
StepSearchBackbone=Qwen2.5-3b-Instruct2026.0117.4————————————
Search-R1Backbone=Qwen2.5-7B-Instruct2026.0117.4————————————
LLM-EmbedderSearch Agent=Our Search Agent, Search Agent Domain=in-domain2026.0117.33————————————
Agentic-RSearch Agent=SimpleDeepSearcher, Search Agent Domain=out-of-domain2026.0117.29————————————
DeepResearcher2026.021726.55————2.5444—————
GSPOBackbone=Qwen2.5-3b2025.0917————————————
AutoRefine-BaseBackbone=Qwen2.5-3B, Retrieval Protocol=w/ Multi-Hop Retrieval2025.1216.9————————————
ZeroSearchBackbone=Qwen2.5-7B-Instruct2026.0116.8————————————
RAGModel Backbone=Qwen3-4B2026.0316.6————————————
SCARLetSearch Agent=Our Search Agent, Search Agent Domain=in-domain2026.0116.38————————————
MultiSearchBackbone=3B-Base2026.0516.325.618.1——————————
BGESearch Agent=Our Search Agent, Search Agent Domain=in-domain2026.0116.21————————————
CoT-SCBackbone=Qwen2.5-14B-Instruct, Method Strategy=Search-based2025.041657.46———————————
REPLUGSearch Agent=SimpleDeepSearcher, Search Agent Domain=out-of-domain2026.0115.8————————————
AutoRefine2026.0515.723.317.5——————————
E5Search Agent=SimpleDeepSearcher, Search Agent Domain=out-of-domain2026.0115.55————————————
Self-Ask + Search Engine2022.1015.227.219.6——————————
GRPOBackbone=Qwen2.5-3b2025.0915.2————————————
TSPOBackbone=Qwen2.5-7b-Instruct2026.0115.1————————————