| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| LongMemEval | ElasticMem | Accuracy76 | 22 | 1mo ago | |
| LoCoMo | ElasticMem | Accuracy84 | 22 | 1mo ago | |
| ATM-Bench Hard | QS Score74.7 | 7 | 4mo ago | ||
| ATM-Bench | QS Score86 | 7 | 4mo ago | ||
| WorldLines Memory QA samples | ObsMem | Judge Score71.3 | 5 | 1mo ago |