Memory-Augmented Question Answering on AdaMem-Bench Explicit Feedback 1.0 (test)
96.4AccuracyFull Context
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Full ContextQA Evaluator=DeepSeek-V4-Flash, Extractor=Extractor-independent2026.06 | 96.4 | — | — | 0 | |
| Ideal MemoryQA Evaluator=DeepSeek-V4-Flash, Extractor=Extractor-independent2026.06 | 96 | 97.9 | 100 | 121 | |
| AdaMemExtraction model=DeepSeek-V4-Flash, QA Evaluator=DeepSeek-V4-Flash2026.06 | 85.2 | 40.1 | 28.9 | 371 | |
| AdaMemExtraction model=Gemini-3.5-Flash, QA Evaluator=DeepSeek-V4-Flash2026.06 | 84.8 | 47.2 | 30.3 | 352 | |
| Mem0Extraction model=Gemini-3.5-Flash, QA Evaluator=DeepSeek-V4-Flash2026.06 | 80.4 | 42.4 | 26.7 | 378 | |
| Mem0Extraction model=DeepSeek-V4-Flash, QA Evaluator=DeepSeek-V4-Flash2026.06 | 76.2 | 35 | 22.3 | 430 |