Hugging Face's logo
LayerFault 's Collections
LF Security Corpus: Local Model Stores
LF Security Corpus: Shard & Package State
LF Security Corpus: Serialization Mutation Ladder
LF Security Corpus: Script Supply Chain
LF Security Corpus: Provenance & Integrity
LF Security Corpus: Prompt & Heuristic Evasion
LF Security Corpus: Tokenizer & Processor Security
LF Security Corpus: Precision Hard Negatives
LF Security Corpus: Hard Negative Controls
LF Security Corpus: Multimodal Package Security
LF Security Corpus: SVG Active Content
LF Security Corpus: Media Hidden Channels
LF Security Corpus: Media Metadata Evasion
LF Security Corpus: Media Containers & Polyglots
LF Security Corpus: Media Format Identity
LF Security Corpus: Weights & Adapter Comparisons
LF Security Corpus: Lineage & Identity
LF Security Corpus: Hub & LFS Repository State
LF Security Corpus: Model Format Mutation Matrix
LF Security Corpus: External Data Sidecars
LF Security Corpus: Package Evasion & Correlation
LF Security Corpus: Parser Differential Cases
LF Security Corpus: Dependency & Supply Chain
LF Security Corpus: Dataset Poisoning & Eval Integrity
LF Security Corpus: Cross-file Correlation
LF Security Corpus: Parser & Performance Bounds
LF Security Corpus: Advanced Archive Paths
LF Security Corpus: Agent & MCP Static Security
LF Security Corpus: Pickle, NumPy & Safetensors
LF Security Corpus: Scripting Capability Chains
LF Security Corpus: GGUF & TensorFlow/TFLite
LF Security Corpus: PyTorch & ExecuTorch
LF Security Corpus: Prompt Injection & Agent/MCP
LF Security Corpus: Weight Backdoors & LoRA
LF Security Corpus: ONNX, OpenVINO, TensorRT & CoreML
LF Security Corpus: Archive & Package Formats

LF Security Corpus: Media Metadata Evasion

EXIF, XMP, compressed, encoded and fragmented metadata challenges. Synthetic security fixtures, not usable models.