List your company

Task-oriented Dialogue on MultiWOZ 2.1 (SR, HR)

51.24Success Rate (SR)VLK-RL
27.590433.730239.8746.0098Apr 25, 2026

Evaluation results

MethodLinks
VLK-RLbackbone=Qwen-14B2026.0451.243.18
VLK-RLbackbone=Qwen-7B2026.0449.363.07
VLK-RLbackbone=GPT-4o-mini2026.0447.133.04
GDP-Zero2026.0441.22.61
CAPID2026.0440.562.63
TransferTOD2026.0439.822.58
GALAXY2026.0438.752.46
ACGOS2026.0434.12.35
PPO2026.0428.52.21