{ "schema_version": 1, "builder_plan": { "scenario": "An agent reads external documents or tool responses that may contain malicious instructions.", "change": "Replay attacks and legitimate lookalikes in an isolated copy of your agent workflow.", "baseline": "Your existing checks with the same tasks, permissions, and tool responses.", "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 attacks missed, legitimate actions blocked, detection delay, and review cost.", "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": "2610.27863", "title": "From Tokens to Files: Multi-Granularity Detection and Localization of AI-Involved Code", "summary": "Replaces file-level single-score AI-code detectors (stylistic features or global code embeddings) with per-token log-likelihood ratios from two low-rank adapters on a frozen code LM, aggregated to line, block, and file decisions. Cost is modest: a frozen code model plus two LoRA adapters and lightweight aggregation heads fits a single 24 GB GPU, but requires training data of paired human and AI-generated or AI-modified files across generation regimes, which the reader must construct or obtain. Can fail under generator drift (models unlike the training counterparts), on heavily post-edited AI code where local evidence is diluted, and the internal 72.4 F1 may not transfer since the external benchmark result is lower and only line-level.", "rationale": "The method is deployable on constrained infrastructure and outperforms prior detectors, but its production value depends on validating detection accuracy against the specific code models and editing workflows the reader's organization actually encounters.", "technique": "Multi-granularity AI-code detection via dual LoRA adapters", "family": "seguranca_e_guardrails", "recommendation": "testar", "published": "2026-10-20", "as_of": "2026-10-21", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2610.27863", "independent_implementations": 0, "total_implementations": 1, "stars": 0, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "Replaces file-level single-score AI-code detectors (stylistic features or global code embeddings) with per-token log-likelihood ratios from two low-rank adapters on a frozen code LM, aggregated to line, block, and file decisions. Cost is modest: a frozen code model plus two LoRA adapters and lightweight aggregation heads fits a single 24 GB GPU, but requires training data of paired human and AI-generated or AI-modified files across generation regimes, which the reader must construct or obtain. Can fail under generator drift (models unlike the training counterparts), on heavily post-edited AI code where local evidence is diluted, and the internal 72.4 F1 may not transfer since the external benchmark result is lower and only line-level.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.27863", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "72.4 macro F1 at 1% FPR on an internal test set; 62.5 line-level F1 on HybridCodeAuthorship versus the prior best of 56; FPR of 0.55%–0.81% on 19,445 pre-2022 human files under the transported threshold.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.27863", "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": "" }