- Using LLMs to recover semantic information (variable names, types, etc) from stripped or obfuscated binaries
- Prompting strategies and context-window management for whole-program reasoning
- Measuring hallucination rates
Please contact Sebastian Schrittwieser.