Text Summarization on Summary (LLM-as-judge)
44.4LLM-as-judge ScoreProbMoE
Evaluation Results
| Method | Links | |
|---|---|---|
| ProbMoEBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 44.4 | |
| DenseMixerBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 41 | |
| ProbMoEBackbone=OLMoE-1B-7B, Top-k=82026.06 | 39.29 | |
| ConventionalBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 39 | |
| DefaultMoEBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 38.4 | |
| Frozen RouterBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 38.29 | |
| DenseMixerBackbone=OLMoE-1B-7B, Top-k=82026.06 | 37.5 | |
| Frozen RouterBackbone=OLMoE-1B-7B, Top-k=82026.06 | 36.05 | |
| ConventionalBackbone=OLMoE-1B-7B, Top-k=82026.06 | 33.7 | |
| Base ModelBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 28.29 | |
| ReMoEBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 25.8 | |
| Base ModelBackbone=OLMoE-1B-7B, Top-k=82026.06 | 7.4 | |
| SparseMixerBackbone=Qwen1.5-MoE-A2.7B, Top-k=42026.06 | 2.1 |