List your company

Single-Hop Question Answering on Trivia

68.1Exact MatchTree-GRPO
20.2632.6845.157.52Sep 25, 2025

Evaluation results

MethodLinks
Tree-GRPOBackbone=Qwen2.5-14b2025.0968.1
GSPOBackbone=Qwen2.5-14b2025.0967.4
GRPOBackbone=Qwen2.5-14b2025.0967.2
GSPOBackbone=Qwen2.5-7b2025.0964.5
ReActBackbone=Qwen2.5-14b2025.0964.2
Search-o1Backbone=Qwen2.5-14b2025.0963.5
Tree-GRPOBackbone=Qwen2.5-7b2025.0963.3
GRPOBackbone=Qwen2.5-7b2025.0961.5
Tree-GRPOBackbone=Llama3.2-3b2025.0959.9
Tree-GRPOBackbone=Qwen2.5-3b2025.0959.7
GSPOBackbone=Qwen2.5-3b2025.0958.8
GRPOBackbone=Llama3.2-3b2025.0958.2
GRPOBackbone=Qwen2.5-3b2025.0958
GSPOBackbone=Llama3.2-3b2025.0957.8
Tree-GRPOBackbone=Qwen2.5-1.5b2025.0957.3
ReActBackbone=Qwen2.5-7b2025.0956.3
Direct InferenceBackbone=Qwen2.5-14b2025.0953.1
GRPOBackbone=Qwen2.5-1.5b2025.0951
GSPOBackbone=Qwen2.5-1.5b2025.0948.9
Search-o1Backbone=Llama3.2-3b2025.0948.4
Search-o1Backbone=Qwen2.5-7b2025.0947.2
Search-o1Backbone=Qwen2.5-3b2025.0944.3
ReActBackbone=Qwen2.5-3b2025.0943.5
ReActBackbone=Llama3.2-3b2025.0942.4
Direct InferenceBackbone=Qwen2.5-7b2025.0940.8
Search-o1Backbone=Qwen2.5-1.5b2025.0930.9
Direct InferenceBackbone=Llama3.2-3b2025.0929.6
Direct InferenceBackbone=Qwen2.5-3b2025.0928.8
Direct InferenceBackbone=Qwen2.5-1.5b2025.0922.4
ReActBackbone=Qwen2.5-1.5b2025.0922.1