List your company

Multi-hop Question Answering on MuSiQue multi-hop (test)

19.4Average @16Dr. MAS
7.4410.54513.6516.755Feb 9, 2026

Evaluation results

MethodLinks
Dr. MASBackbone=Qwen2.5-7B, Agent Configuration=Multi-Agent & LLM Non-Sharing2026.0219.434.9
GRPOBackbone=Qwen2.5-7B, Agent Configuration=Single-Agent2026.0217.834.6
Dr. MASBackbone=Qwen2.5-7B, Agent Configuration=Multi-Agent & LLM Sharing2026.0216.732.1
GRPOBackbone=Qwen2.5-7B, Agent Configuration=Multi-Agent & LLM Sharing2026.0215.231.7
Dr. MASBackbone=Qwen2.5-3B, Agent Configuration=Multi-Agent & LLM Non-Sharing2026.0210.426.1
Dr. MASBackbone=Qwen2.5-3B, Agent Configuration=Multi-Agent & LLM Sharing2026.0210.225.8
GRPOBackbone=Qwen2.5-3B, Agent Configuration=Multi-Agent & LLM Sharing2026.029.126.5
GRPOBackbone=Qwen2.5-3B, Agent Configuration=Multi-Agent & LLM Non-Sharing2026.028.619.2
GRPOBackbone=Qwen2.5-7B, Agent Configuration=Multi-Agent & LLM Non-Sharing2026.028.318.1
GRPOBackbone=Qwen2.5-3B, Agent Configuration=Single-Agent2026.027.914.6