About
Research & Development
Making AI Accountable: Falsifying and Quantifying Explanations in Deep Learning
TLDR: This paper introduces a framework to move from "explainable" to "explained" AI in medical imaging. It defines what constitutes a good explanation for...
Research & Development
Unlocking AI Vulnerabilities: A New Approach to Multimodal Model Jailbreaking
TLDR: A new research paper introduces Balanced Structural Decomposition (BSD), a novel method for jailbreaking Multimodal Large Language Models (MLLMs). BSD effectively bypasses AI...
Research & Development
Navigating AI Ethics: A Global Survey Reveals Varied Practices Across Development Teams
TLDR: A survey of 414 AI professionals across 43 countries reveals varying familiarity and application of AI ethics principles and regulations. While AI is...
Research & Development
Advancing Mammography Reporting with AI: Introducing the AMRG Framework
TLDR: The AMRG framework, built on the medical-domain specialized MedGemma-4B-it VLM and using efficient LoRA fine-tuning, is the first end-to-end system for generating narrative...
Research & Development
Cowpox: A New Defense Against Digital Viruses in AI Multi-Agent Systems
TLDR: COWPOX is a novel defense mechanism designed to protect Vision Language Model (VLM)-based multi-agent systems from infectious jailbreak attacks. It works by deploying...
Research & Development
Designing Understandable AI: A Human-Centered Approach to Explanations
TLDR: This research paper proposes a human-centered framework for designing Explainable AI (XAI) systems, moving beyond purely technical methods. It introduces a 'Who, What,...