Science Question Answering on ARC (Easy/Challenge Breakdown)
92.11ARC Easy AccuracyLD-MoLE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LD-MoLEModel=Qwen3-1.7B, TP=4.23 %2025.09 | 92.11 | 78.67 | |
| ReMoLEModel=Qwen3-1.7B, TP=4.12 %2025.09 | 91.75 | 79.6 | |
| ReMoLEModel=Llama3.1-8B, TP=2.17 %2025.09 | 91.61 | 82 | |
| LD-MoLEModel=Llama3.1-8B, TP=2.33 %2025.09 | 91.61 | 83.67 | |
| LD-MoLEModel=Llama3.2-3B, TP=3.28 %2025.09 | 89.47 | 74.58 | |
| ReMoLEModel=Llama3.2-3B, TP=3.11 %2025.09 | 89.3 | 75.25 | |
| MoLA(8888)Model=Qwen3-1.7B, TP=4.12 %2025.09 | 88.6 | 76.59 | |
| MoLA(2468)Model=Qwen3-1.7B, TP=2.39 %2025.09 | 88.42 | 76.92 | |
| MoLA(2468)Model=Llama3.1-8B, TP=1.26 %2025.09 | 87.37 | 77.9 | |
| MoLA(8888)Model=Llama3.1-8B, TP=2.17 %2025.09 | 86.32 | 76.79 | |
| MoLA(2468)Model=Llama3.2-3B, TP=1.80 %2025.09 | 83.86 | 71.91 | |
| MoLA(8888)Model=Llama3.2-3B, TP=3.11 %2025.09 | 83.51 | 71.57 | |
| RPOChain-of-thought=false, Few-shots=false, Backbone=Gemma2024.05 | 79.1 | 49.8 | |
| zephyr-gemma-7bChain-of-thought=false, Few-shots=false2024.05 | 77.6 | 48.6 | |
| DPOChain-of-thought=false, Few-shots=false, Backbone=Gemma2024.05 | 75.7 | 50 | |
| zephyr-7b-gemma-sftChain-of-thought=false, Few-shots=false, Backbone=Gemma2024.05 | 75 | 45.8 |