List your company

Multi-hop Question Answering on HotpotQA (pass@1, pass@2)

65.9pass@1 Success RateEXG
25.96436.33246.757.068May 18, 2026

Evaluation results

MethodLinks
EXGModel=Qwen-Max2026.0565.968.4
EXG-SEModel=Qwen-Max2026.0565.770.7
EXG-SE-RevModel=Qwen-Max2026.0565.570.3
EXG-ReflexionModel=Qwen-Max2026.0565.168.6
SE-Agent-RevModel=Qwen-Max2026.0562.168
EXGModel=Qwen-Plus2026.0561.864.2
ReflexionModel=Qwen-Max2026.0561.668.6
EXG-SEModel=Qwen-Plus2026.0561.368.1
EXG-SE-RevModel=Qwen-Plus2026.0560.868
SE-AgentModel=Qwen-Max2026.0560.766.9
SE-AgentModel=Qwen-Plus2026.0560.367
EXG-ReflexionModel=Qwen-Plus2026.0560.167.4
ReflexionModel=Qwen-Plus2026.0559.766.2
SE-Agent-RevModel=Qwen-Plus2026.0558.766.9
EXG-ReflexionModel=Qwen3-8B2026.0546.255.7
EXGModel=Qwen3-8B2026.054650
EXG-SE-RevModel=Qwen3-8B2026.0545.850.9
EXG-SEModel=Qwen3-8B2026.0544.551.6
ReflexionModel=Qwen3-8B2026.0528.349.4
SE-AgentModel=Qwen3-8B2026.0528.341
SE-Agent-RevModel=Qwen3-8B2026.0527.534.4