Planning on Countdown (Countdown metric)
19.4CountdownLLaDA + Dream
Evaluation Results
| Method | Links | |
|---|---|---|
| LLaDA + DreamEnsemble Type=Intermediate-generation Ensemble, Scoring Function=Entropy2026.06 | 19.4 | |
| LLaDA + DreamEnsemble Type=Post-generation Ensemble2026.06 | 19.2 | |
| LLaDA + DreamEnsemble Type=Intermediate-generation Ensemble, Scoring Function=Token change count2026.06 | 18.8 | |
| LLaDA + DreamEnsemble Type=Intermediate-generation Ensemble, Scoring Function=Top-1 probability2026.06 | 18.8 | |
| LLaDA + DreamEnsemble Type=Intermediate-generation Ensemble, Scoring Function=Probability margin2026.06 | 18.6 | |
| DreamEnsemble Type=Single Model2026.06 | 16.4 | |
| LLaDAEnsemble Type=Single Model2026.06 | 13.4 |