List your company

Question Answering on 2WikiMultiHopQA (F1, LM, Time)

43Exact MatchTIPS
17.10423.82730.5537.273May 10, 2026May 13, 2026May 16, 2026May 19, 2026May 22, 2026May 25, 2026May 29, 2026

Evaluation results

MethodLinks
TIPSBackbone=Qwen2.5-7B Instruct2026.0543—————
All-pool PortfolioModel=Llama-3.1-70B-Instruct, k=4, l=32026.0541.87—————
All-pool PortfolioModel=Gemma-3-27B-It, k=4, l=32026.0541.42—————
All-pool PortfolioModel=Llama-3.1-70B-Instruct, k=4, l=22026.0541.39—————
PiCABackbone=Qwen2.5-3B Instruct2026.0540.8—————
All-pool PortfolioModel=Gemma-3-27B-It, k=4, l=22026.0540.52—————
Search-R1Backbone=Qwen2.5-7B Instruct2026.0540.1—————
PiCABackbone=Qwen2.5-7B Instruct2026.0540.1—————
StepSearchBackbone=Qwen2.5-7B Instruct2026.0536.6—————
Best Vendi RetrieverModel=Gemma-3-27B-It2026.0535.62—————
Best DS RetrieverModel=Gemma-3-27B-It2026.0535.44—————
RAG-GDBackbone=Llama-3.1-8B, Retriever=BM25, K=52026.0534.15—40.04———
Best GraphDense RetrieverModel=Gemma-3-27B-It2026.0534.07—————
RAG-GDBackbone=Llama-3.1-8B, Retriever=E5, K=52026.0533.94—39.83———
Base adapterBackbone=Llama-3.1-8B, Retriever=BM252026.0533.66—39.55———
Base adapterBackbone=Llama-3.1-8B, Retriever=E52026.0533.45—39.45———
StepSearchBackbone=Qwen2.5-3B Instruct2026.0532—————
Search-R1Backbone=Qwen2.5-3B Instruct2026.0531.9—————
ZeroSearchBackbone=Qwen2.5-7B Instruct2026.0530.9—————
Vendi-Portfolio (MPNet)Model=Llama-3.1-70B-Instruct, k=5, l=22026.0530.8—————
RAG-GDBackbone=Qwen-2.5-7B, Retriever=E5, K=52026.0529.67—35.47———
TIPSBackbone=Qwen2.5-3B Instruct2026.0529.3—————
Nearest-neighbor retrieval (MPNet)Model=Llama-3.1-70B-Instruct2026.0529.21—————
Vendi-RAG Adaptive (T=20) (MPNet)Model=Llama-3.1-70B-Instruct, T=202026.0529—————
Vendi-RAG Adaptive (T=10) (MPNet)Model=Llama-3.1-70B-Instruct, T=102026.0528.9—————
RAG-GDBackbone=Qwen-2.5-7B, Retriever=BM25, K=52026.0528.86—34.48———
Base adapterBackbone=Qwen-2.5-7B, Retriever=BM252026.0528.22—34.13———
Base adapterBackbone=Qwen-2.5-7B, Retriever=E52026.0527.89—33.97———
Vanilla RAGBackbone=Qwen-2.5-7B, Retriever=BM252026.0527.87—33.24———
Query OnlyBackbone=Llama-3.1-8B, Retriever=None2026.0526.39—31.04———
Vanilla RAGBackbone=Llama-3.1-8B, Retriever=BM252026.0526.07—31.82———
Vendi-RAG Adaptive (T=10) (MPNet)Model=Gemma-3-27B-It, T=102026.0525.9—————
Vendi-RAG Adaptive (T=20) (MPNet)Model=Gemma-3-27B-It, T=202026.0525.6—————
Vanilla RAGBackbone=Qwen-2.5-7B, Retriever=E52026.0525.48—31.43———
Best Vendi RetrieverModel=Llama-3.1-70B-Instruct2026.0524.45—————
Best DS RetrieverModel=Llama-3.1-70B-Instruct2026.0524.43—————
Nearest-neighbor retrieval (MPNet)Model=Gemma-3-27B-It2026.0524.08—————
Query OnlyBackbone=Qwen-2.5-7B, Retriever=None2026.0523.91—28.12———
ZeroSearchBackbone=Qwen2.5-3B Instruct2026.0523.9—————
Best GraphDense RetrieverModel=Llama-3.1-70B-Instruct2026.0523.87—————
Vanilla RAGBackbone=Llama-3.1-8B, Retriever=E52026.0523.44—29.46———
Vendi-Portfolio (MPNet)Model=Gemma-3-27B-It, k=5, l=22026.0523.41—————
RAGBackbone=Qwen2.5-7B Instruct2026.0523.1—————
MT-PPOBackbone=Qwen2.5-7B Instruct2026.0522.8—————
No retrievalModel=Gemma-3-27B-It2026.0522.64—————
RAGBackbone=Qwen2.5-3B Instruct2026.0522.1—————
MT-PPOBackbone=Qwen2.5-3B Instruct2026.0521.4—————
Search-o1Backbone=Qwen2.5-3B Instruct2026.0520.7—————
No retrievalModel=Llama-3.1-70B-Instruct2026.0519.2—————
Search-o1Backbone=Qwen2.5-7B Instruct2026.0518.1—————
ChameleonModel=Qwen-Plus2025.01—56.6859.1860.121.737.53
ChameleonModel=LLaMA-3.1-8B2025.01—40.5243.4940.271.8426.71
GoldRouterModel=Qwen-Plus2025.01—70.1172.3776.331.3630.93
GoldRouterModel=LLaMA-3.1-8B2025.01—49.7352.9952.071.5820.57
HiPRAGBackbone=Qwen2.5-7B2025.10———42.8——
MT-PPOBackbone=Qwen2.5-3B Instruct2026.05——26.1———
MT-PPOBackbone=Qwen2.5-7B Instruct2026.05——28.6———
PiCABackbone=Qwen2.5-3B Instruct2026.05——47.2———
PiCABackbone=Qwen2.5-7B Instruct2026.05——46.6———
RAGBackbone=Qwen2.5-3B Instruct2026.05——31.6———
RAGBackbone=Qwen2.5-7B Instruct2026.05——22.6———
ReSeekBackbone=Qwen2.5-7B2025.10———38.2——
RIRSModel=Qwen-Plus2025.01—61.1463.7165.781.7339.76
RIRSModel=LLaMA-3.1-8B2025.01—45.6549.145.652.0222.26
RIRSModel=Mixed, Note=RAG agents use LLaMA-3.1-8B, other modules use Qwen-Plus2025.01—59.4861.9462.061.8539.37
RouterDCModel=Qwen-Plus2025.01—50.9753.8852.121.6745.61
RouterDCModel=LLaMA-3.1-8B2025.01—34.4837.3530.591.6726.58
Search-o1Backbone=Qwen2.5-3B Instruct2026.05——30.9———
Search-o1Backbone=Qwen2.5-7B Instruct2026.05——28.9———
Search-R1Backbone=Qwen2.5-3B Instruct2026.05——35.2———
Search-R1Backbone=Qwen2.5-7B Instruct2026.05——46.5———
StepSearchBackbone=Qwen2.5-3B Instruct2026.05——38.5———
StepSearchBackbone=Qwen2.5-7B Instruct2026.05——43.1———
TIPSBackbone=Qwen2.5-3B Instruct2026.05——35.1———
TIPSBackbone=Qwen2.5-7B Instruct2026.05——50.3———
ZeroSearchBackbone=Qwen2.5-3B Instruct2026.05——28.8———
ZeroSearchBackbone=Qwen2.5-7B Instruct2026.05——37———