EvoMemBench Model Card
这是可索引项目报告证据页：它保留 EvoMemBench Model Card 的源材料入口、机制线索和限制提醒；正文仍需 reader/editor 与 academic public-copy review 后才能当作最终结论引用。
EvoMemBench Model Card
| Field | Value |
|---|---|
| Repository | DSAIL-Memory/EvoMemBench |
| Category | Self-Evolving Agent Memory Benchmark |
| Stars / forks snapshot | 0 / 0 |
| Language | Markdown |
| License | Apache-2.0 |
| Raw capture | raw-github/dsail-memory_evomembench.md |
| Updated by | hourly public metadata update, 2026-05-25 |
1. Role in Self Evolve
EvoMemBench 是 2026-05 arXiv 论文配套的 agent memory benchmark，按 in-episode/cross-episode 与 knowledge/execution 两条轴评估记忆机制是否支持自演化 agent。
2. Working Principle
agent tasks -> memory mechanism variants -> in/cross-episode benchmark axes -> self-evolving memory evidence
3. Evidence Path
web GitHub page observed public repository with 2 commits, Apache-2.0 license, 0 stars and 0 forks, README says Coming soon; arXiv 2605.18421 states the code is available at this repository and frames EvoMemBench as agent memory benchmarking from a self-evolving perspective. Shell GitHub API access remained blocked by DNS and local gh auth was invalid, so this card treats the current snapshot as web-observed rather than API-verified.
4. Teaching Use
Use this card to explain Self-Evolving Agent Memory Benchmark in the raw -> classification -> project card -> site/report pipeline. The key question is whether memory/context evidence improves downstream agent behavior, or only improves retrieval on a local benchmark.
5. Limits
当前未克隆源码，未运行 benchmark、SDK examples、skill install flows、memory experiments、agent evolution loops 或 production deployments；star/fork/commit 快照来自公开 GitHub 页面文本或可见页面片段。