List your company

Multi-Hop Question Answering on Bamb

54.4Exact Match (EM)Tree-GRPO
0.3214.3628.442.44Sep 25, 2025

Evaluation results

MethodLinks
Tree-GRPOBackbone=Qwen2.5-14b2025.0954.4
GRPOBackbone=Qwen2.5-14b2025.0953.6
GSPOBackbone=Qwen2.5-14b2025.0952.8
Tree-GRPOBackbone=Llama3.2-3b2025.0946.4
GSPOBackbone=Qwen2.5-7b2025.0944
Tree-GRPOBackbone=Qwen2.5-7b2025.0944
Tree-GRPOBackbone=Qwen2.5-3b2025.0943.2
GRPOBackbone=Qwen2.5-7b2025.0943.2
ReActBackbone=Qwen2.5-14b2025.0943.2
Search-o1Backbone=Qwen2.5-14b2025.0941.6
GRPOBackbone=Qwen2.5-3b2025.0936.8
GSPOBackbone=Qwen2.5-3b2025.0936.8
Search-o1Backbone=Llama3.2-3b2025.0932
GRPOBackbone=Llama3.2-3b2025.0932
GSPOBackbone=Llama3.2-3b2025.0932
Search-o1Backbone=Qwen2.5-7b2025.0932
Search-o1Backbone=Qwen2.5-3b2025.0929.6
ReActBackbone=Qwen2.5-7b2025.0928.8
ReActBackbone=Llama3.2-3b2025.0923.2
ReActBackbone=Qwen2.5-3b2025.0920
Direct InferenceBackbone=Qwen2.5-14b2025.0916
Tree-GRPOBackbone=Qwen2.5-1.5b2025.0913.6
Search-o1Backbone=Qwen2.5-1.5b2025.0913
Direct InferenceBackbone=Qwen2.5-7b2025.0912
ReActBackbone=Qwen2.5-1.5b2025.0911.2
Direct InferenceBackbone=Qwen2.5-1.5b2025.098
Direct InferenceBackbone=Llama3.2-3b2025.098
GSPOBackbone=Qwen2.5-1.5b2025.094.8
GRPOBackbone=Qwen2.5-1.5b2025.094
Direct InferenceBackbone=Qwen2.5-3b2025.092.4