Learning Agent Routing From Early Experience
Yimin Wang Jiahao Qiu Xuan Qi Xinzhe Juan Jingzhe Shi, [0.25em] Zelin Zhao Hongru Wang Shilong Liu Mengdi Wang, [0.6em] AI Lab, Princeton University University of Michigan · 2026
Read Learning Agent Routing From Early Experience by Yimin Wang Jiahao Qiu Xuan Qi Xinzhe Juan Jingzhe Shi, [0.25em] Zelin Zhao Hongru Wang Shilong Liu Mengdi Wang, [0.6em] AI Lab, Princeton University University of Michigan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-07180.txt.
Description
Yimin Wang and eight co-authors route questions between a quick language model reply and a full agent run. Agents handle hard reasoning well but are slow and costly, while many queries sit within a plain model's reach. Their BoundaryRouter decides which path each query needs, without any training.
It builds a small memory from early experience, running both systems on a shared seed set and retrieving similar cases to guide choices. On their new RouteBench, covering paraphrased and out-of-domain cases too, the router cut time sharply versus the agent. It also improved performance over direct model replies and beat prompt-based routing.
Catalog record
| CREATOR | Yimin Wang Jiahao Qiu Xuan Qi Xinzhe Juan Jingzhe Shi, [0.25em] Zelin Zhao Hongru Wang Shilong Liu Mengdi Wang, [0.6em] AI Lab, Princeton University University of Michigan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.07180