{"data":{"id":"61ccf367-38d6-4d94-8342-ce4e0b98a16e","title":"Compile-Time Backdoor Neutralization via LLM-Guided Localized Edits","summary":"CBN is a defense framework against compile-time backdoors in deep learning compilers such as TVM, where a trigger effect is re-materialized during compilation. It performs localized rewriting at the compiler-IR level, guided by runtime probes and LLM-proposed plans from a finite edit vocabulary. Across seven ONNX-Zoo models and five commercial LLMs, it reduces mean DCL-BD attack success from 30.1% to 1.9% with 2.0% mean inference-latency overhead.","solution":"CBN: a defense framework for compilation-induced backdoors that performs localized rewriting at the compiler-IR level, guided by runtime probes and LLM-proposed plans drawn from a finite edit vocabulary, with a legality-aware schema, a budgeted-transformation validator, and a probe-based feedback loop that repairs broken plans.","labels":["security","research"],"sourceUrl":"http://ieeexplore.ieee.org/document/11793824","publishedAt":"2026-11-12T05:12:32.000Z","cveId":null,"cweIds":null,"cvssScore":null,"cvssSeverity":null,"severity":"info","attackType":["other"],"issueType":"research","affectedPackages":null,"affectedPackageNames":null,"affectedPackageRefs":null,"affectedVendors":[],"affectedVendorsRaw":["TVM","ONNX Runtime","StableHLO/XLA","ONNX-Zoo models","commercial LLMs"],"classifierModel":"claude-haiku-5-5","classifierPromptVersion":"v4","summaryPromptVersion":"v2","headline":null,"headlinePromptVersion":null,"cvssVector":null,"attackVector":null,"attackComplexity":null,"privilegesRequired":null,"userInteraction":null,"exploitMaturity":null,"epssScore":null,"epssCheckedAt":null,"kevDateAdded":null,"advisoryAliases":null,"affectedPackagesSource":null,"affectedPackagesCheckedAt":null,"patchAvailable":null,"disclosureDate":"2026-11-12T05:12:32.000Z","fixReleasedAt":null,"daysToFix":null,"fixCheckedAt":null,"capecIds":null,"crossRefCount":0,"attackSophistication":"advanced","impactType":["integrity"],"aiComponentTargeted":"framework","llmSpecific":false,"classifierConfidence":0.9,"researchCategory":"peer_reviewed","atlasIds":null}}