Key Information Retrieval on Needle-in-a-Haystack 32K context
98.2Accuracy
MEAP
Updated 1mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2025.02 | 98.2 | |
2026.03 | 93.85 | |
2026.03 | 91.15 | |
2026.03 | 90.29 | |
2026.03 | 87.23 | |
2025.02 | 87.1 | |
2026.03 | 86.82 | |
2025.02 | 85.8 | |
2026.03 | 80.29 | |
2025.02 | 80.2 | |
2026.03 | 77.33 | |
2026.03 | 77.2 | |
2026.03 | 75.84 | |
2026.03 | 71.88 | |
2026.03 | 68.21 | |
2025.02 | 65.9 | |
2025.02 | 52.8 | |
2026.03 | 50.88 | |
2026.03 | 46.81 |