Date Prediction on TARA (test)
11.6EM@1NewsRECON
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| NewsRECONEvaluation Setup=Simplified, Constraint=Choice limited to test set labels2026.01 | 11.6 | 30.2 | 47.4 | 43.5 | 58.2 | |
| CLIP+Evaluation Setup=Simplified, Constraint=Choice limited to test set labels2026.01 | 8.1 | 23.3 | 44.6 | 38.2 | 55.2 | |
| Qwen2.5VL-7BEvaluation Setup=Open-ended, Evidence Source=NewsRECON retrieved articles2026.01 | 5.8 | — | 46.8 | 47.6 | 62.1 | |
| InternVL3-8BEvaluation Setup=Open-ended, Evidence Source=NewsRECON retrieved articles2026.01 | 4.4 | — | 45.4 | 44.1 | 60.4 | |
| NewsRECONEvaluation Setup=Open-ended2026.01 | 3.7 | 11.1 | 43.5 | 42.2 | 59.1 | |
| Qwen2.5VL-7BEvaluation Setup=Open-ended, Prompting Strategy=celebrity2026.01 | 3.3 | — | 40.1 | 41.7 | 59.7 | |
| Qwen2.5VL-7BEvaluation Setup=Open-ended, Prompting Strategy=default2026.01 | 3.2 | — | 39.4 | 40.7 | 59.3 | |
| InternVL3-8BEvaluation Setup=Open-ended, Prompting Strategy=celebrity2026.01 | 2.2 | — | 42.7 | 39.2 | 57.5 | |
| InternVL3-8BEvaluation Setup=Open-ended, Prompting Strategy=default2026.01 | 2 | — | 42 | 37.5 | 56.3 | |
| Qwen2.5VL-7BEvaluation Setup=Open-ended, Prompting Strategy=GETReason2026.01 | 2 | — | 38.6 | 33.6 | 58.5 | |
| InternVL3-8BEvaluation Setup=Open-ended, Prompting Strategy=GETReason2026.01 | 0.6 | — | 27.1 | 15.1 | 33.8 |