MEME: Multi-entity & Evolving Memory Evaluation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910213353766912 |
|---|---|
| author | Jung, Seokwon Rubinstein, Alexander Uselis, Arnas Yun, Sangdoo Oh, Seong Joon |
| author_facet | Jung, Seokwon Rubinstein, Alexander Uselis, Arnas Yun, Sangdoo Oh, Seong Joon |
| contents | LLM-based agents increasingly operate in persistent environments where they must store, update, and reason over information across many sessions. While prior benchmarks evaluate only single-entity updates, MEME defines six tasks spanning the full space defined by the multi-entity and evolving axes, including three not scored by prior work: Cascade and Absence (dependency reasoning) and Deletion (post-removal state). Evaluating six memory systems spanning three memory paradigms on 100 controlled episodes, we find that all systems collapse on dependency reasoning under the default configuration (Cascade: 3%, Absence: 1% in average accuracy) despite adequate static retrieval performance. Prompt optimization, deeper retrieval, reduced filler noise, and most stronger LLMs fail to close this gap. Only a file-based agent paired with Claude Opus 4.7 as its internal LLM partially closes the gap, but at ~70x the baseline cost, indicating closure currently depends on configurations that are not practical at scale. Code and data are available on the project page: https://seokwonjung-jay.github.io/meme-eval/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_12477 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MEME: Multi-entity & Evolving Memory Evaluation Jung, Seokwon Rubinstein, Alexander Uselis, Arnas Yun, Sangdoo Oh, Seong Joon Machine Learning Computation and Language LLM-based agents increasingly operate in persistent environments where they must store, update, and reason over information across many sessions. While prior benchmarks evaluate only single-entity updates, MEME defines six tasks spanning the full space defined by the multi-entity and evolving axes, including three not scored by prior work: Cascade and Absence (dependency reasoning) and Deletion (post-removal state). Evaluating six memory systems spanning three memory paradigms on 100 controlled episodes, we find that all systems collapse on dependency reasoning under the default configuration (Cascade: 3%, Absence: 1% in average accuracy) despite adequate static retrieval performance. Prompt optimization, deeper retrieval, reduced filler noise, and most stronger LLMs fail to close this gap. Only a file-based agent paired with Claude Opus 4.7 as its internal LLM partially closes the gap, but at ~70x the baseline cost, indicating closure currently depends on configurations that are not practical at scale. Code and data are available on the project page: https://seokwonjung-jay.github.io/meme-eval/. |
| title | MEME: Multi-entity & Evolving Memory Evaluation |
| topic | Machine Learning Computation and Language |
| url | https://arxiv.org/abs/2605.12477 |