List your company

Question Answering on HotpotQA (EM, F1, EFR)

58.4EMPruneRAG
8.06421.13234.247.268Jan 16, 2026

Evaluation results

MethodLinks
PruneRAGBackbone=Qwen3-32B, Time=22212026.0158.46624.6
PruneRAGBackbone=Llama-3.1-70B-Instruct, Time=14322026.0158.26629.2
RAGBackbone=Qwen3-32B, Time=13942026.0154.262.621.9
RAGBackbone=Llama-3.1-70B-Instruct, Time=6032026.0150.657.229.7
ConTRGenBackbone=Llama-3.1-70B-Instruct, Time=227992026.0150.657.730.7
ConTRGenBackbone=Qwen3-32B, Time=169212026.0149.858.736.4
Search-o1Backbone=Llama-3.1-70B-Instruct, Time=30572026.0146.252.640.2
Search-o1Backbone=Qwen3-32B, Time=50472026.0142.449.443.6
RAG-StarBackbone=Llama-3.1-70B-Instruct, Time=432392026.0141.650.949.6
ReactBackbone=Llama-3.1-70B-Instruct, Time=28242026.0141.246.641.8
RAG-StarBackbone=Qwen3-32B, Time=525272026.0138.446.949.3
ProbTreeBackbone=Llama-3.1-70B-Instruct, Time=25832026.0136.444.650.5
ReactBackbone=Qwen3-32B, Time=79272026.0134.440.435.5
Self-RAGBackbone=Llama-3.1-70B-Instruct, Time=52942026.0130.637.559.1
Self-RAGBackbone=Qwen3-32B, Time=52942026.0130.637.559.1
ProbTreeBackbone=Qwen3-32B, Time=25042026.0125.638.159
VanillaBackbone=Llama-3.1-70B-Instruct, Time=5152026.012530.8—
VanillaBackbone=Qwen3-32B, Time=34592026.0121.225.4—
MemoRAGBackbone=Llama-3.1-70B-Instruct, Time=10002026.01101010
MemoRAGBackbone=Qwen3-32B, Time=10002026.01101010