Completed Pairs Hide Capped Failures: A ReVerPi Case Study of Selective Context Projection· 完成对的隐藏上限失败：ReVerPi 案例研究
Context projection replaces older tool observations with compact, addressable excerpts, reducing repeated input while potentially adding evidence-retrieval turns. We study this trade-off in ReVerPi, a Pi extension with archived observations and matched full/projected continuations. In an 86-run source-reading campaign with 641 model requests, the 15 completed pairs show identical success: 12/15 per arm. Twelve further boundary runs stop, with the runner suppressing the companion whenever the first arm fails to complete. Restoring all 27 boundary runs bounds projected-minus-full success between $-$9 and +1 tasks. One omitted, selector-chosen projected continuation successfully retrieves archive text yet exhausts twelve requests; its full counterpart answers in three. The eleven jointly corr
研究上下文投影在 ReVerPi 中的应用与影响。
- 核心方法
- 通过 ReVerPi 框架，比较带有存档观察的上下文投影与完整上下文的模型请求成功率。
- 适合谁读
- 研究者
- 要解决的问题
- 上下文投影技术可能引入额外的证据检索轮次，如何评估其成本与收益。
- 关键实验
- 进行了 86 次源代码阅读活动，包含 641 次模型请求，其中 15 次完成对显示相同成功率，边界运行重启后评估投影与完整上下文差异。
- 主要贡献
- 发现完成对的隐藏上限失败现象，并提出边界运行重启策略。
- 意义与局限
- 为上下文投影技术提供了实证支持，但也指出了其可能的局限性。