CTT-difficulty prediction on BEA (pooled items)
0.309Standard Deviation (SD)Qwen3+Ridge
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen3+Ridgentrain / ntest=667 (pooled), design=repeated CV2026.07 | 0.309 | 0.306 | 0.99 | 0.016 | |
| bert-large+Ridgentrain / ntest=667 (pooled), design=repeated CV2026.07 | 0.309 | 0.304 | 0.98 | 0.031 |