Metraipolis › archive › dc_0e6aee0570db

SAGE: Symbolic Action-Gating and Editing for LLM Task Planners — arXiv:2609.34268

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.34268
License (as archived)
CC0-1.0
Archived
2026-10-07T21:02:19.027Z
SHA-256 of the archived text
4bb6c78c88d28f260e9d418401292b48536fe7c3ef6a4cf5b1b77773bb3995ec
Stable URL of this page
https://metraipolis.com/documento/dc_0e6aee0570db

How to cite

SAGE: Symbolic Action-Gating and Editing for LLM Task Planners — arXiv:2609.34268 — arXiv (en), https://arxiv.org/abs/2609.34268 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0e6aee0570db (Metraipolis archive dc_0e6aee0570db, sha256 4bb6c78c88d28f260e9d418401292b48536fe7c3ef6a4cf5b1b77773bb3995ec)
BibTeX
@misc{dc0e6aee0570db,
  title = {SAGE: Symbolic Action-Gating and Editing for LLM Task Planners — arXiv:2609.34268},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34268},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0e6aee0570db; sha256 4bb6c78c88d28f260e9d418401292b48536fe7c3ef6a4cf5b1b77773bb3995ec},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34268. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language models (LLMs) are now the default cognitive core of embodied household agents, yet the plans they emit are rarely checked against a grounded model of the environment before execution, and the task-success they report is often measured on benchmarks so saturated that no method can be separated from another. We present SAGE (Symbolic Action-Gating and Editing), a single-LLM planner built from two lightweight mechanisms: a domain-agnostic symbolic gate (~250 lines of Python, zero tokens, $O(|\pi|)$) that blocks precondition-violating actions with typed reasons as a runtime safety monitor, and a local edit that regenerates only the failed sub-goal's suffix, keeping completed and untouched work intact; a hybrid seed+live memory store supports cold-start coverage. We evaluate under a leak-free protocol (leave-one-out retrieval) over five open-weight models and a 75-task AI2-THOR benchmark. On the standard benchmark goal-completeness saturates (52% of instances trivially solved) and SAGE ties strong hierarchical baselines. On a harder, method-agnostic multi-goal composition, SAGE's completeness lead re-emerges large (+0.06 to +0.23 across four models). Under injected mid-execution failures, SAGE recovers as reliably as whole-plan replanners at 2.4-3.3x fewer LLM calls. As a verify-before-execute gate, the symbolic monitor blocks unsafe actions before actuation and raises simulator-reported step-success for every planner tested (up to +0.11), a signal the verifier never sees (non-circular). Because the gate calls no model (0.008 ms/plan), it is a safety layer that runs essentially free on the edge: SAGE planning reproduces its quality on a Jetson AGX Orin, where small-model verification helps most. We release the benchmark, the leak-free protocol, the recovery and safety-gate harnesses, and a verifier-portability study (auto-induced on ALFWorld, 0.89 held-out).

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.34268 — 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 2026-10-07T20:29:02.913Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).