6 papers
Not All 4-bit Quantizers Are Equal: Deployment-Time Mitigation of PII Leakage in Fine-Tuned Small Language Models
Cristhian Kapelinski, Diego Kreutz
Organizations fine-tune small language models on private data and then compress them to 4 bits for resource-efficient deployment. We show that the compression method also affects p…
Vulnerabilities, Secrets and Misconfiguration in the Highest-Exposure Docker Hub Images
Cristhian Kapelinski, Beatriz Machado, Diego Kreutz
Docker Hub is the registry underneath most container deployments, and a flaw in a widely reused base image is inherited by every image built on it. Prior ecosystem-scale measuremen…
Decomposing Memorization Reduction in Privacy-Preserving Fine-Tuning of SLMs for CSIRTs
Cristhian Kapelinski, Diego Kreutz
CSIRTs increasingly fine tune language models on vulnerability scan records, but these records expose internal network topology and create privacy risks under regulations such as G…
AnonShield: Scalable On-Premise Pseudonymization for CSIRT Vulnerability Data
Cristhian Kapelinski, Douglas Lautert, Beatriz Machado +2
We present AnonShield, a high-throughput, on-premise pseudonymization system that combines GPU-accelerated NER, streaming processing, caching, and schema-aware configuration. Evalu…
Structured Extraction of Vulnerabilities in OpenVAS and Tenable WAS Reports Using LLMs
Beatriz Machado, Douglas Lautert, Cristhian Kapelinski +1
This paper proposes an automated LLM-based method to extract and structure vulnerabilities from OpenVAS and Tenable WAS scanner reports, converting unstructured data into a standar…
AnonLFI 2.0: Extensible Architecture for PII Pseudonymization in CSIRTs with OCR and Technical Recognizers
Cristhian Kapelinski, Douglas Lautert, Beatriz Machado +1
This work presents AnonLFI 2.0, a modular pseudonymization framework for CSIRTs that uses HMAC SHA256 to generate strong and reversible pseudonyms, preserves XML and JSON structure…