{ "schema_version": 1, "builder_plan": { "scenario": "A multi-step product task fails because the agent loses track of dependencies.", "change": "Try the proposed planning step on tasks with checkable intermediate outcomes.", "baseline": "The same agent solving the same tasks without the new planning step.", "steps": [ "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.", "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.", "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.", "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test." ], "measure": "Task completion, unnecessary steps, planning tokens, response time, and total cost.", "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation." }, "builder_review": null, "arxiv_id": "2610.18874", "title": "ChunkAct: Action Chunking for Agentic Robotics", "summary": "ChunkAct replaces per-primitive LLM reasoning over full interaction history with offline-compiled chunks of chained primitives that execute autonomously and return control to the planner only on failure, making LLM cost scale with decisions rather than primitive count. The cost is an offline compilation step that converts the agent's memory into chunks, plus loss of per-step planner oversight, since execution becomes memory-free between decisions. It can fail when a sub-goal's routine does not match the compiled chunk, when failure detection within a chunk is unreliable, or when tasks require mid-primitive replanning that the chunk structure cannot express.", "rationale": "The chunking pattern is directly transferable to any tool-calling agent workload where repeated LLM invocations execute routine sequences, but the reported gains come from a robotics benchmark, so adoption in a general production agent stack requires validating chunk reliability and failure detection on the reader's own tasks.", "technique": "Action chunking for LLM-driven control loops", "family": "planejamento_e_decomposicao", "recommendation": "testar", "published": "2026-10-13", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2610.18874", "independent_implementations": 0, "total_implementations": 0, "stars": 0, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "ChunkAct replaces per-primitive LLM reasoning over full interaction history with offline-compiled chunks of chained primitives that execute autonomously and return control to the planner only on failure, making LLM cost scale with decisions rather than primitive count. The cost is an offline compilation step that converts the agent's memory into chunks, plus loss of per-step planner oversight, since execution becomes memory-free between decisions. It can fail when a sub-goal's routine does not match the compiled chunk, when failure detection within a chunk is unreliable, or when tasks require mid-primitive replanning that the chunk structure cannot express.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.18874", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "On LIBERO-Pro, ChunkAct improves task success by 5.0 points over the prior state of the art while requiring 2.9x fewer LLM calls, 4.4x fewer input tokens, and 2.3x fewer output tokens.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.18874", "source_page": null, "source_excerpt": "" } ], "exposure_map": [ { "dimension": "quality", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compute", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "latency", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "operations", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compatibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "security", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "data_and_training", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "reproducibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" } ], "risks": [], "minimum_test": [], "open_questions": [], "independent_tests": [], "lab_reproductions": [], "report_available": false, "validation_tier": "unknown", "evidence_tier": "unknown", "infrastructure_basis": "unknown", "equations": [], "equations_status": "not_fetched", "core_kind": "", "equations_fetched_at": "", "equations_source_url": "", "equations_source_sha256": "", "exposures_reviewed_at": "", "exposures_source_sha256": "" }