{ "schema_version": 1, "builder_plan": { "scenario": "An agent looks good in demos, but you cannot tell which product tasks it fails.", "change": "Use the study's evaluation idea to inspect a small set of tasks with human-checked answers.", "baseline": "Your current evaluator and a manually reviewed reference set.", "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": "Agreement with human review, missed failures, false alarms, and evaluation cost per 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": "2609.25647", "title": "Testing-Driven Reliability Audit of Trajectory-Based Early Outcome Prediction for LLM Agents: Target-Specific Calibration Transfer Persists Within a Single Benchmark", "summary": "The approach replaces running every agent evaluation trajectory to completion with a trajectory-based early-outcome predictor that terminates runs once the outcome is confidently predictable, cutting evaluation compute and API spend. The cost is an additional frozen dual-head prediction pipeline plus mandatory per-target calibration auditing (leave-one-agent-out audits, oracle prior correction, resampling and threshold robustness checks) before trusting any early termination decision. What can fail is silent target-specific miscalibration: particular agent/head combinations showed large persistent corrected gaps, and the failure did not replicate across benchmarks, so a predictor calibrated on one agent or benchmark cannot be assumed safe for a new production agent.", "rationale": "Early-exit prediction is directly deployable on constrained infrastructure and can reduce evaluation API cost, but the paper demonstrates that calibration transfer can fail for specific target agents, so adoption is justified only after running the described per-target calibration audit on the reader's own agent and workload.", "technique": "Calibrated early-exit prediction for agent evaluation", "family": "avaliacao_de_agente", "recommendation": "testar", "published": "2026-09-22", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.25647", "independent_implementations": 0, "total_implementations": 0, "stars": 0, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "The approach replaces running every agent evaluation trajectory to completion with a trajectory-based early-outcome predictor that terminates runs once the outcome is confidently predictable, cutting evaluation compute and API spend. The cost is an additional frozen dual-head prediction pipeline plus mandatory per-target calibration auditing (leave-one-agent-out audits, oracle prior correction, resampling and threshold robustness checks) before trusting any early termination decision. What can fail is silent target-specific miscalibration: particular agent/head combinations showed large persistent corrected gaps, and the failure did not replicate across benchmarks, so a predictor calibrated on one agent or benchmark cannot be assumed safe for a new production agent.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.25647", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "No quantified savings factor is reported; the paper instead quantifies calibration risk, finding persistent corrected calibration gaps of 0.1377 (gpt-5-mini/SUCCESS head) and 0.1107 (claude-opus-4.6/FAILURE head) under frozen leave-one-agent-out transfer, while broad pairwise heterogeneity was not supported (median gaps 0.0180 and 0.0385).", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.25647", "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": "" }