Science Question Answering on Materials (Mat.)
79.3Mean@16SC-SDPO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SC-SDPOBackbone=Qwen3-8B, Alpha=0.52026.05 | 79.3 | 80.3 | |
| SDPOBackbone=Qwen3-8B2026.05 | 78 | 78.6 | |
| SDPOBackbone=Qwen3-8B, Filtering Strategy=Hard Filter2026.05 | 77.3 | 78.1 | |
| SRPOBackbone=Qwen3-8B2026.05 | 77.2 | 80.3 | |
| SC-SDPOBackbone=Qwen3-8B, Alpha=12026.05 | 77.1 | 78.3 | |
| SDPOBackbone=Qwen3-8B, Weighting Strategy=PACED2026.05 | 75.4 | 75.8 | |
| GRPOBackbone=Qwen3-8B, Normalization=None2026.05 | 73.3 | 74.5 | |
| GRPOBackbone=Qwen3-8B2026.05 | 69.8 | 69.9 | |
| Qwen3-8BBackbone=Qwen3-8B2026.05 | 59.7 | 59.6 |