Symbolic Reasoning on Countdown
49.61AccuracyFDM-A
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FDM-ALLDM Model=LLaDA-MoE2025.12 | 49.61 | 10.14 | |
| ProbabilityLLDM Model=LLaDA-MoE, Threshold (T)=1282025.12 | 44.53 | 8.8 | |
| MarginLLDM Model=LLaDA-MoE, Threshold (T)=1282025.12 | 42.58 | 8.33 | |
| EBLLDM Model=LLaDA-MoE2025.12 | 42.19 | 9.48 | |
| EntropyLLDM Model=LLaDA-MoE, Threshold (T)=1282025.12 | 35.94 | 8.15 | |
| FDM-ALLDM Model=MMaDA-MixCoT2025.12 | 26.95 | 62.3 | |
| WINOLLDM Model=LLaDA-MoE2025.12 | 25 | 8.23 | |
| FDM-ALLDM Model=LLaDA2025.12 | 21.48 | 21.98 | |
| ProbabilityLLDM Model=LLaDA-1.5, Threshold (T)=1282025.12 | 21.09 | 22.01 | |
| FDM-ALLDM Model=LLaDA-1.52025.12 | 21.09 | 22.29 | |
| WINOLLDM Model=MMaDA-MixCoT2025.12 | 20.7 | 52.39 | |
| MarginLLDM Model=LLaDA, Threshold (T)=1282025.12 | 20.31 | 20.65 | |
| MarginLLDM Model=LLaDA-1.5, Threshold (T)=1282025.12 | 19.92 | 21.76 | |
| EBLLDM Model=LLaDA-1.52025.12 | 19.92 | 18.79 | |
| ProbabilityLLDM Model=LLaDA, Threshold (T)=1282025.12 | 19.53 | 21.19 | |
| EntropyLLDM Model=LLaDA, Threshold (T)=1282025.12 | 19.53 | 20.62 | |
| EBLLDM Model=LLaDA2025.12 | 19.2 | 18.68 | |
| WINOLLDM Model=LLaDA2025.12 | 19.14 | 20.52 | |
| WINOLLDM Model=LLaDA-1.52025.12 | 19.14 | 20.38 | |
| EntropyLLDM Model=LLaDA-1.5, Threshold (T)=1282025.12 | 18.75 | 20.93 | |
| EBLLDM Model=MMaDA-MixCoT2025.12 | 18.75 | 48.51 | |
| ProbabilityLLDM Model=MMaDA-MixCoT, Threshold (T)=1282025.12 | 12.01 | 21.97 | |
| MarginLLDM Model=MMaDA-MixCoT, Threshold (T)=1282025.12 | 8.59 | 21.57 | |
| EntropyLLDM Model=MMaDA-MixCoT, Threshold (T)=1282025.12 | 8.59 | 20.52 |