3 papers
cs.CR2026
The Geometry of Harmfulness in Multi-Turn Attacks
Yelyzaveta, Husieva, Lauren Alvarez
Large language models (LLMs) remain vulnerable to adversarial attacks that circumvent safety alignment to elicit harmful outputs. It remains unclear how harmfulness and refusal rep…
cs.CL2025
Mapping Semantic & Syntactic Relationships with Geometric Rotation
Michael Freenor, Lauren Alvarez
Understanding how language and embedding models encode semantic relationships is fundamental to model interpretability. While early word embeddings exhibited intuitive vector arith…
cs.CR2025
Prompt Optimization and Evaluation for LLM Automated Red Teaming
Michael Freenor, Lauren Alvarez, Milton Leal +8
Applications that use Large Language Models (LLMs) are becoming widespread, making the identification of system vulnerabilities increasingly important. Automated Red Teaming accele…