| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Long-term Memory Question Answering | ATM-Bench 1.0 (test) | QS77.8 | 15 | |
| Memory Question Answering | ATM-Bench Hard | QS Score74.7 | 7 | |
| Memory Question Answering | ATM-Bench | QS Score86 | 7 | |
| Long-horizon Memory Question Answering | ATM-Bench subset431 | EM69.61 | 3 |