{
  "schema_version": 1,
  "builder_plan": {
    "scenario": "A narrow, repeated product task may not need your largest model.",
    "change": "Check whether a released student model fits the task before generating training data.",
    "baseline": "Your current model and an off-the-shelf small model 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": "Student quality, serving cost, teacher calls, training cost, and failures on unusual inputs.",
    "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": "2610.29167",
  "title": "TeachBack: Learning to Teach via Bidirectional Preference Optimization for On-Policy Self-Distillation",
  "summary": "TeachBack replaces standard on-policy self-distillation where the Teacher has static privileged access to the solution with a bidirectional preference-learning framework that shapes the Teacher's generation via reference-length control and prompt variation. It requires generating paired privileged and non-privileged trajectories at each rollout refresh and optimizing a pairwise preference objective, which increases training compute and implementation complexity compared to standard OPSD. The method can fail if the privileged Teacher trajectories, despite accuracy, omit the exploration and self-correction behaviors necessary for the Student to learn robust problem-solving without privileged information.",
  "rationale": "While the method targets distillation—a relevant area for constrained infrastructure—the requirement for on-policy bi-directional preference optimization with paired trajectory generation and custom reward modeling exceeds the limited compute and implementation budget of a small team without clusters.",
  "technique": "Bidirectional preference optimization for self-distillation",
  "family": "destilacao",
  "recommendation": "observar",
  "published": "2026-10-21",
  "as_of": "2026-10-25",
  "editorial_status": "indexed",
  "source_url": "https://arxiv.org/abs/2610.29167",
  "independent_implementations": 0,
  "total_implementations": 1,
  "stars": 0,
  "citations": 0,
  "claims": [
    {
      "claim_id": "claim-01",
      "statement": "TeachBack replaces standard on-policy self-distillation where the Teacher has static privileged access to the solution with a bidirectional preference-learning framework that shapes the Teacher's generation via reference-length control and prompt variation. It requires generating paired privileged and non-privileged trajectories at each rollout refresh and optimizing a pairwise preference objective, which increases training compute and implementation complexity compared to standard OPSD. The method can fail if the privileged Teacher trajectories, despite accuracy, omit the exploration and self-correction behaviors necessary for the Student to learn robust problem-solving without privileged information.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.29167",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "claim_id": "claim-02",
      "statement": "The abstract states that TeachBack consistently improves Student performance over existing OPSD methods and related baselines on mathematical and logical reasoning tasks, but provides no specific quantified improvement in the provided text.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.29167",
      "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": ""
}