List your company

General QA on PopQA

52Exact Match (EM)IRCoT
-1.5612.34526.2540.155Oct 1, 2025Oct 31, 2025Nov 30, 2025Dec 30, 2025Jan 29, 2026Feb 28, 2026Mar 31, 2026

Evaluation results

MethodLinks
IRCoTBackbone=Qwen2.5-72B-Instruct2026.035264.5
ReSeekBackbone=Qwen2.5-7b-Instruct2025.1050.1—
ZeroSearchBackbone=Qwen2.5-7b-Instruct2025.1048.8—
ZeroSearchBackbone=Qwen2.5-3b-Instruct2025.1044.8—
ReSeekBackbone=Qwen2.5-3b-Instruct2025.1043.4—
HippoRAGBackbone=Qwen2.5-72B-Instruct2026.0341.149.7
Search-R1Backbone=Qwen2.5-7b-Instruct2025.1039.7—
RAGBackbone=Qwen2.5-7b-Instruct2025.1039.2—
RAGBackbone=Qwen2.5-3b-Instruct2025.1038.7—
Search-R1Backbone=Qwen2.5-3b-Instruct2025.1037.8—
TaSR-RAGBackbone=Qwen2.5-72B-Instruct2026.0337.743.2
GraphRAG-RouterSystem=Our Method2026.0336.8—
GraphRAGBackbone=Qwen2.5-72B-Instruct2026.0336.745.8
StructRAGBackbone=Qwen2.5-72B-Instruct2026.0336.545
TaSR-RAGBackbone=Qwen2.5-7B-Instruct2026.0336.543.9
HyperGraphRAGBackbone=Qwen2.5-72B-Instruct2026.033645.9
Router-R1System=RL-based Agent2026.0335.1—
RAGBackbone=Qwen2.5-72B-Instruct2026.0333.241.2
CoTSystem=Basic LLM2026.0332.7—
GraphSearchSystem=Training-free Agent2026.0332.4—
StructRAGBackbone=Qwen2.5-7B-Instruct2026.0331.440.1
Search-o1System=Training-free Agent2026.0330.2—
IRCoTBackbone=Qwen2.5-7B-Instruct2026.0330.160.2
GraphRAGSystem=RAG-Based2026.0329.1—
CoT ReasoningBackbone=GPT-4o-mini2025.1128.1332.66
Prompt-R1Backbone=GPT-4o-mini2025.1128.1333.77
Graph-R1System=RL-based Agent2026.0327.4—
GEPAOptimization Framework=APO, Backbone=GPT-4o-mini2025.1127.3431.18
HyperGraphRAGSystem=RAG-Based2026.0327.2—
LinearRAGSystem=RAG-Based2026.0326.9—
HippoRAGBackbone=Qwen2.5-7B-Instruct2026.0326.837
HippoRAG2System=RAG-Based2026.0326.6—
Search-o1Backbone=Qwen2.5-3b-Instruct2025.1026.2—
RAPTORSystem=RAG-Based2026.0325.8—
BaselineBackbone=GPT-4o-mini2025.1125.7830.75
Search-R1System=RL-based Agent2026.0325.6—
CoTBackbone=Qwen2.5-72B-Instruct2026.0324.430.3
Vanilla RAGSystem=RAG-Based2026.0324.2—
OPROOptimization Framework=APO, Backbone=GPT-4o-mini2025.1123.4430.14
R1Backbone=Qwen2.5-7b-Instruct2025.1019.9—
Direct InferenceBackbone=Qwen2.5-72B-Instruct2026.0319.724.4
TextGradOptimization Framework=APO, Backbone=GPT-4o-mini2025.1119.5328.28
RAGBackbone=Qwen2.5-7B-Instruct2026.0319.244.3
R1Backbone=Qwen2.5-3b-Instruct2025.1017.1—
Direct InferenceBackbone=Qwen2.5-7B-Instruct2026.031416.4
Direct InferenceBackbone=Qwen2.5-7b-Instruct2025.1014—
Search-o1Backbone=Qwen2.5-7b-Instruct2025.1013.1—
GraphRAGBackbone=Qwen2.5-7B-Instruct2026.0312.827.4
Direct InferSystem=Basic LLM2026.0311.7—
Direct InferenceBackbone=Qwen2.5-3b-Instruct2025.1010.8—
SFTBackbone=Qwen3-4B2025.117.819.53
GRPOBackbone=Qwen3-4B2025.117.8112.27
BaselineBackbone=Qwen3-4B2025.117.039.97
CoT ReasoningBackbone=Qwen3-4B2025.117.039.58
CoTBackbone=Qwen2.5-7B-Instruct2026.035.416.4
CoTBackbone=Qwen2.5-7b-Instruct2025.105.4—
HypergraphRAGBackbone=Qwen2.5-7B-Instruct2026.030.81.6
CoTBackbone=Qwen2.5-3b-Instruct2025.100.5—