Decision Making on Hiring Domain
90.2Top-1 AccuracyDECISIVE
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| DECISIVEDialogue=true2026.04 | 90.2 | 97.6 | 0.958 | 0.947 | 1.9 | |
| LLM-CoT (B2)Dialogue=false2026.04 | 87.8 | 97.8 | 0.949 | 0.935 | — | |
| LLM-Direct (B1)Dialogue=false2026.04 | 87.2 | 97.2 | 0.942 | 0.931 | — | |
| Active Pref. InferenceDialogue=true2026.04 | 32 | 56.6 | 0.749 | 0.554 | 7.5 | |
| LLM-Free Dialogue (B4)Dialogue=true2026.04 | 31.6 | 40.6 | 0.416 | 0.472 | 2.4 | |
| LLM-Structured Dialogue (B3)Dialogue=true2026.04 | 11.4 | 21.8 | 0.229 | 0.303 | 6.9 | |
| DeLLMaDialogue=false2026.04 | 9.8 | 20 | 0.667 | 0.288 | — | |
| DecisionFlowDialogue=false2026.04 | 9.2 | 20.4 | 0.665 | 0.292 | — |