List your company

Multi-turn dialogue on ShareGPT, JDDC, and MedDG Aggregated

89.77SRavgMCTS Router
50.04260.35670.6780.984Apr 14, 2026

Evaluation results

MethodLinks
MCTS RouterCandidate LLM Set=Qwen Series Set2026.0489.77
MCTS RouterCandidate LLM Set=Llama Series Set2026.0486.53
Greedy RouterCandidate LLM Set=Qwen Series Set2026.0483.94
DialRouterCandidate LLM Set=Qwen Series Set2026.0483.31
Greedy RouterCandidate LLM Set=Llama Series Set2026.0479.7
DialRouterCandidate LLM Set=Llama Series Set2026.0478.88
AvengersCandidate LLM Set=Qwen Series Set2026.0478
KNN RouterCandidate LLM Set=Qwen Series Set2026.0477.93
Qwen3-14BCandidate LLM Set=Qwen Series Set2026.0477.64
RouterDCCandidate LLM Set=Qwen Series Set2026.0477.47
Prompt LLMCandidate LLM Set=Qwen Series Set2026.0477.36
MF RouterCandidate LLM Set=Qwen Series Set2026.0477.14
Qwen3-8BCandidate LLM Set=Qwen Series Set2026.0476.92
RandomCandidate LLM Set=Qwen Series Set2026.0476.52
Qwen3-4BCandidate LLM Set=Qwen Series Set2026.0475.41
KNN RouterCandidate LLM Set=Llama Series Set2026.0473.7
RouterDCCandidate LLM Set=Llama Series Set2026.0473.6
AvengersCandidate LLM Set=Llama Series Set2026.0473.5
MF RouterCandidate LLM Set=Llama Series Set2026.0472.75
Llama3.1-8B-InstructCandidate LLM Set=Llama Series Set2026.0472.6
Prompt LLMCandidate LLM Set=Llama Series Set2026.0472.34
RandomCandidate LLM Set=Llama Series Set2026.0468.75
Llama3.2-3B-InstructCandidate LLM Set=Llama Series Set2026.0468.4
Llama3.2-1B-InstructCandidate LLM Set=Llama Series Set2026.0451.57