{ "schema_version": 1, "builder_plan": { "scenario": "Start with one repeated product task that has a measurable failure or cost.", "change": "Read the methods and identify one change you can isolate in that workflow.", "baseline": "Your current implementation on the same held-out tasks.", "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": "Successful outcomes, response time, human rework, and total cost per successful task.", "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": "2609.37359", "title": "Encore: Few-Shot Agentic Discovery of Manipulation Strategies", "summary": "ENCORE replaces training-data-style imitation learning and hand-crafted robot policy specification with a pipeline in which demonstrations are distilled into keyframe and trajectory packs that a coding agent reads to write and iteratively refine a policy program against a fixed perception/action API. It costs a small set of demonstrations per task (five on the real robot), development rollouts, and the deterministic distillation infrastructure, plus robotics hardware and a simulation benchmark that the reader's setup does not include. It fails when demonstrations are absent or instructions leave the goal unstated (no success on RoboDojo without demonstrations), and frozen programs can overfit to development rollouts since the sealed evaluation signal is hidden.", "rationale": "The method addresses robot manipulation policy discovery, which is outside the reader's production scope of LLM inference, agents, and retrieval work on a 24 GB GPU or API budget.", "technique": "Few-shot demonstration-conditioned code generation for robot policies", "family": "outro", "recommendation": "nao_aplica", "published": "2026-09-29", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.37359", "independent_implementations": 1, "total_implementations": 2, "stars": 224, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "ENCORE replaces training-data-style imitation learning and hand-crafted robot policy specification with a pipeline in which demonstrations are distilled into keyframe and trajectory packs that a coding agent reads to write and iteratively refine a policy program against a fixed perception/action API. It costs a small set of demonstrations per task (five on the real robot), development rollouts, and the deterministic distillation infrastructure, plus robotics hardware and a simulation benchmark that the reader's setup does not include. It fails when demonstrations are absent or instructions leave the goal unstated (no success on RoboDojo without demonstrations), and frozen programs can overfit to development rollouts since the sealed evaluation signal is hidden.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.37359", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "Frozen agent-written policies reach 96.3% on LIBERO-PRO versus 89.3% for the strongest prior agentic system with the same language model; first-attempt programs succeed on roughly half of perturbed tasks with demonstrations.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.37359", "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": "" }