From ReAct to AFlow: The Workflow Turn in LLM Agents
Published:
Blog posts
2026
Indirect Prompt Injection in LLM Applications and Agents: Threat Models, Benchmarks, and Defense Mechanisms
Published:
Indirect prompt injection (IPI) refers to attacks where malicious instructions are embedded in untrusted external content (e.g., web pages, documents, tool responses) and are later consumed by an LLM application alongside the user’s intended instruction. Early work demonstrated that this threat is not hypothetical: real-world LLM-integrated systems can be manipulated through injected content, enabling outcomes such as data exfiltration and unintended tool misue.
2025
Adding Geospatial Features to Your Machine Learning Models
Published:
Introduction
How Natural Language Processing Can Map Human Stories
Published:
When disaster strikes, people don’t just wait for the news - they post.
Why Geography Still Matters in Our High-Tech World
Published: