Hi, I'm Aflah, a research software engineer at the Max Planck Institute for Software Systems. My primary focus is on advancing our understanding of large language models (LLMs), evaluating their capabilities, and developing AI powered co-pilots to support researchers. I'm currently working on optimizing pre-training and inference for LLMs as well as understanding the challenges posed by the widespread use of AI agents. Previously, I’ve worked on projects aimed at reducing hate speech on social media and other applications under NLP for social good.
Open to researcher/research engineer/backend engineer roles
- Paper Reading & Discussion: Implicit meta-learning may lead LMs to trust more reliable sources
- Paper Reading & Discussion: When Does Metadata Con. (NOT) Work for LM Pre-Training? A Study with CFG
- Paper Reading & Discussion: URLs Help, Topics Guide: Understanding Metadata Utility in LLM Training
- Paper Reading & Discussion: Metadata Conditioning Accelerates Language Model Pre-training
- Paper Reading & Discussion: Open-Reasoner-Zero: An OS Approach to Scaling Up RL on the Base Model