{ "schema_version": 1, "builder_plan": { "scenario": "You already need task-specific training and want to reduce its memory or compute cost.", "change": "Change only the training technique on a small controlled run after checking prerequisites.", "baseline": "Your current training recipe with the same model, data split, and evaluation.", "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": "Held-out quality, peak memory, training time, and full compute cost to the target quality.", "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": "2610.06204", "title": "Do Small Language Models Learn to Negotiate? A Controlled Scaling Study of RL-Trained Sellers", "summary": "The method replaces prompting an off-the-shelf base model (or paying for a frontier API) as a negotiating seller with GRPO fine-tuning of a 4.5B-12B open model on a programmatic bargaining reward. Costs are RL training infrastructure (a 4.5B seller fits one 48 GB GPU for serving; GRPO training requires more memory and careful rollout generation), plus the engineering of a verifiable reward and evaluation harness against held-out buyer models. It can fail through buyer-family overfitting (the 2.3B high-rate arm improved mainly against the buyer sharing its training family), through a base model already matching frontier sellers so RL adds little, and through learning-rate sensitivity that can make a small model appear untrainable when it is merely undertuned.", "rationale": "For a constrained team building a negotiating agent, RL on a 4.5B-12B open model is feasible at small scale, but the paper's single-run arms, architecture confounds, and buyer-family overfitting mean the quality gain must be validated on the reader's own counterparties before adoption.", "technique": "GRPO RL training of small seller models", "family": "treino_eficiente", "recommendation": "testar", "published": "2026-10-05", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2610.06204", "independent_implementations": 0, "total_implementations": 1, "stars": 121, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "The method replaces prompting an off-the-shelf base model (or paying for a frontier API) as a negotiating seller with GRPO fine-tuning of a 4.5B-12B open model on a programmatic bargaining reward. Costs are RL training infrastructure (a 4.5B seller fits one 48 GB GPU for serving; GRPO training requires more memory and careful rollout generation), plus the engineering of a verifiable reward and evaluation harness against held-out buyer models. It can fail through buyer-family overfitting (the 2.3B high-rate arm improved mainly against the buyer sharing its training family), through a base model already matching frontier sellers so RL adds little, and through learning-rate sensitivity that can make a small model appear untrainable when it is merely undertuned.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.06204", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "RL gain over base rises from +0.001 (2.3B) to +0.078 (31B) at a shared 1e-6 learning rate; tripling the rate adds +0.032 to +0.081 at small sizes, and a 12B seller trained at the tripled rate scores above two frontier-model sellers.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.06204", "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": "" }