Metraipolis › archive › dc_7f4a83dfd2af

Credit Where It Matters: Dependency-Aware Policy Optimization for Terminal Agents — arXiv:2610.03634

Source: arXiv
Original: https://arxiv.org/abs/2610.03634
License (as archived): CC0-1.0
Archived: 2026-10-05T22:49:45.742Z
SHA-256 of the archived text: cc6a5bf9bca8d19cdbda2d2861c03dd8f91e74e77af5192a0bd9c4e0c6622af8
Stable URL of this page: https://metraipolis.com/documento/dc_7f4a83dfd2af

How to cite

Credit Where It Matters: Dependency-Aware Policy Optimization for Terminal Agents — arXiv:2610.03634 — arXiv, https://arxiv.org/abs/2610.03634 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7f4a83dfd2af (Metraipolis archive dc_7f4a83dfd2af, sha256 cc6a5bf9bca8d19cdbda2d2861c03dd8f91e74e77af5192a0bd9c4e0c6622af8)

BibTeX

@misc{dc7f4a83dfd2af,
  title = {Credit Where It Matters: Dependency-Aware Policy Optimization for Terminal Agents — arXiv:2610.03634},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.03634},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7f4a83dfd2af; sha256 cc6a5bf9bca8d19cdbda2d2861c03dd8f91e74e77af5192a0bd9c4e0c6622af8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.03634. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Terminal-using agents benefit from reinforcement learning (RL) in coding, debugging, and other multi-step terminal tasks. In these tasks, later commands often depend on information or intermediate results produced by earlier commands. However, existing trajectory-level and step-level credit assignment methods do not explicitly trace the read-write dependencies through which commands affect the final outcome. Consequently, training signals could still be assigned to irrelevant operations, weakening learning from relevant steps. In this paper, we propose Dependency-Aware Group Policy Optimization (DepGPO), which uses execution dependencies between commands to guide credit assignment for terminal agents. Specifically, we construct a command dependency graph from execution traces and trace backward from the resources inspected by the task verifier. We then assign credit to relevant writes and their supporting reads along these paths, and use it to redistribute trajectory advantages across steps. Extensive comparative experiments and ablation studies demonstrate that DepGPO improves task performance and training stability on complex terminal tasks.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.03634 — 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 2026-10-05T22:45:55.714Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).