{ "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": "2609.39679", "title": "SE-ADD: Self-Evolving Audio Deepfake Detection with Mistake-Driven Supervision", "summary": "Replaces static, label-only fine-tuning of audio-language-model deepfake detectors with iterative LoRA updates where misclassified samples receive extra supervision from the model's own self-generated forensic cues. Costs are modest in memory (LoRA adapters fit on a 24 GB GPU) but the iterative verdict-and-cue regeneration loop adds training complexity and latency compared to a single fine-tune. It can fail if self-generated forensic cues are wrong, since supervising a model on its own errors risks reinforcing confident mistakes, and gains are reported only for two specific ALMs on the authors' evolving-attack benchmark.", "rationale": "LoRA-based adaptation is feasible on constrained hardware and the EER reductions are large, but the self-supervision loop's reliability must be validated on the reader's own attack distribution before production use.", "technique": "Mistake-driven LoRA self-adaptation for audio deepfake detection", "family": "treino_eficiente", "recommendation": "testar", "published": "2026-09-30", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.39679", "independent_implementations": 0, "total_implementations": 0, "stars": 0, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "Replaces static, label-only fine-tuning of audio-language-model deepfake detectors with iterative LoRA updates where misclassified samples receive extra supervision from the model's own self-generated forensic cues. Costs are modest in memory (LoRA adapters fit on a 24 GB GPU) but the iterative verdict-and-cue regeneration loop adds training complexity and latency compared to a single fine-tune. It can fail if self-generated forensic cues are wrong, since supervising a model on its own errors risks reinforcing confident mistakes, and gains are reported only for two specific ALMs on the authors' evolving-attack benchmark.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.39679", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "Reduces EER from 36.72% to 7.52% on Qwen2-Audio (4.9x relative reduction) and from 19.93% to 3.97% on MOSS-Audio (5.0x) against unseen spoofing attacks.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.39679", "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": "" }