Fact-driven Question Answering on HotpotQA (F1, EM)
60.78F1 Scorew/t BoT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| w/t BoTModel=Qwen2.5-1.5b-instruct, Geometric Component=Barycentric Transport2026.05 | 60.78 | 47.27 | |
| w/t CDModel=Qwen2.5-1.5b-instruct, Geometric Component=Cosine Dispersion2026.05 | 59.97 | 46.8 | |
| w/t KLEModel=Qwen2.5-1.5b-instruct2026.05 | 57.62 | 44.31 | |
| w/t SEModel=Qwen2.5-1.5b-instruct2026.05 | 57.33 | 44.18 | |
| CAREModel=Qwen2.5-1.5b-instruct2026.05 | 57.03 | 43.75 | |
| GRPOModel=Qwen2.5-1.5b-instruct2026.05 | 50 | 37.06 |