02/2025: arXiv preprinted. Position piece on how episodic memory can advance research on LLM agents.
12/2024: Scientific Data, published. The 'Naturalistic Free Recall' Dataset: four stories, hundreds of participants, and high-fidelity transcriptions.
10/2024: arXiv preprinted. Report on our Sequence Order Recall Task (SORT), a benchmark to evaluate long-term memory behavior in LLMs. We release data and code.
09/2024: IEEE High Performance Extreme Computing, presented. Outstanding paper award for Monocoder: Domain-specific code language model for HPC Codes and Tasks.
Research Experience
Recently left my research scientist role at Intel Labs to pursue other projects, both in machine learning and cognitive neuroscience. Previously, I spent 5 years working in AI/ML in the Brain-Inspired Computing Lab at Intel Labs, most recently on vector similarity search, RAG, and LLM code generation.
Background
Interests: At the intersection of computer science, neuroscience, and cognitive science; particularly in the computational underpinnings of intelligent behavior, especially in the use of memory and attention.