Note: body below is original English text extracted from arXiv abs / Atom. Do not treat this file as a translation.
arXiv:2609.05019 · published 2026-09-04
Abstract
Agents tend to optimize, select, or constrain execution structures before decisive runtime outcomes are observed. However, such pre-execution commitment creates an orchestration bottleneck: when intermediate evidence invalidates the pending continuation, agents must either execute stale steps or replan broadly, compounding errors, wasting computation, and discarding progress. We thus propose Trace-grounded Route Orchestration via Validation and Editing (TROVE), which revises only what runtime evidence invalidates. Offline, TROVE distills evaluated workflow-search traces into atomic and composite skills and an outcome-conditioned transition graph, preserving stable fragments while exposing outcome-dependent decisions. Online, it treats a planned route as provisional: after committing one top-level skill, the controller retains a valid continuation, inserts a trace-supported local response, or replaces only the invalid suffix. Evaluation across code-generation, question-answering, and math reasoning benchmarks with different LLM backbones show that TROVE delivers a stronger quality-efficiency trade-off than existing baselines of dataset-level optimization, query-level architecture selection, and graph-constrained scheduling. Quality gains are largest when outcomes change the appropriate continuation, whereas early termination yields substantial efficiency gains on near-saturated tasks. Ablations further show that composite skills capture most offline benefits, insertion enables local correction, and suffix replacement primarily improves efficiency. These findings establish selective route editing as a general principle for adaptive agent orchestration.
Key claims (verbatim-leaning English extract)
- Pre-execution commitment creates orchestration bottleneck when intermediate evidence invalidates continuation.
- Offline: distill workflow-search traces into atomic/composite skills + outcome-conditioned transition graph.
- Online: planned route is provisional — retain valid continuation, insert local response, or replace only invalid suffix.
- Selective route editing beats dataset-level optimization, query-level architecture selection, and graph-constrained scheduling on quality-efficiency.
Remainder
Full original English text: see html_url / source_url / pdf_url in frontmatter. Do not invent missing sections from this partial extract.