Metraipolis › archive › dc_3f167bc81f13

Separate Field Scoring Protects Unused Tools When Agents Pool Tool-Use Experience — arXiv:2610.30867

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.30867
License (as archived)
CC0-1.0
Archived
2026-10-25T08:34:41.601Z
SHA-256 of the archived text
ed29e3add3c33ae659d36f8323fcd9c5d0559473af5c900c94b15761cf6dca79
Stable URL of this page
https://metraipolis.com/documento/dc_3f167bc81f13

How to cite

Separate Field Scoring Protects Unused Tools When Agents Pool Tool-Use Experience — arXiv:2610.30867 — arXiv (en), https://arxiv.org/abs/2610.30867 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3f167bc81f13 (Metraipolis archive dc_3f167bc81f13, sha256 ed29e3add3c33ae659d36f8323fcd9c5d0559473af5c900c94b15761cf6dca79)
BibTeX
@misc{dc3f167bc81f13,
  title = {Separate Field Scoring Protects Unused Tools When Agents Pool Tool-Use Experience — arXiv:2610.30867},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.30867},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3f167bc81f13; sha256 ed29e3add3c33ae659d36f8323fcd9c5d0559473af5c900c94b15761cf6dca79},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.30867. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agents that call tools from a shared catalog can pool what they learn about which tool serves which request. We measure what pooled usage examples do to tool retrieval, reporting separately the held-out tools, for which no agent has usage, and all other tools. Adding usage examples to the index as extra entries for used tools raises Hit@5 (whether a gold tool is in the top five) for those tools and lowers it for held-out tools, by a large margin on two of three catalogs and with both embedders we tested. Giving used tools the same number of equally long entries filled with other tools' text lowers held-out Hit@5 about as much, but gives little or none of the gain. We call our method of scoring usage as its own field SYNAPSE. On StableToolBench, the held-out gold tool appears in its top five more often than with a router that reserves slots for usage (0.785 against 0.683), and this held-out lead holds on all three catalogs with both embedders. This protection persisted when we replaced each tool's usage with other tools' text, and a constant score offset for tools with usage gives similar held-out Hit@5 in most settings. The gain for tools with usage, in contrast, needs usage about the right tool. The router keeps more of the gain on labeled queries on TOUCAN with both embedders and on ToolRet with one, so the better choice depends on the workload. Splitting the usage among ten simulated agents shows that pooling helps exactly as much as giving one agent the same records.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Software Engineering (cs.SE) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.30867 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-25T08:31:13.197Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).