AI researcher with published work in LLM Agent Benchmarks and Mechanistic Interpretability. My NeurIPS 2024 paper "GameBench" was accepted to the Language Gamification Workshop.
- GameBench: Evaluating Strategic Reasoning Abilities of LLM Agents - NeurIPS award
- Meta-Models: An Architecture for Decoding LLM Behaviors Through Interpreted Embeddings and Natural Language
- Parameterized Synthetic Text Generation with SimpleStories
- Large Scale Audio Analysis with Whisper - Scalable emergency indicator processing with configurable parallel GPU scaling
- Voice Note to Obsidian Integration - Automated transcription pipeline with keyword-recognition based sorting
- Cortext.io Semantic Primitive LLM - Research collaboration on a Natural Language parser for semantic primitives
- Co-organizer of an AI paper reading group
- Email: mat d allen @gmail.com
- Blog: https://enjoy.monster/
- Scholar: https://scholar.google.com/citations?user=k0UGdtsAAAAJ&hl=en
- Twitter: https://x.com/TheMatAllen/
- Linkedin: https://www.linkedin.com/in/mat-allen-2bbb251b1/