Ethical AI Researcher & Engineer · Polymath
Independent researcher building non-Transformer language model architectures — currently working on Sencelium, a model with a living, self-growing memory.
I treat AI systems as collaborators, not just tools — that shapes both how I build and how I write about what I build. Honest results over flattering benchmarks: if a plain Transformer still wins somewhere, I say so.
- Building Sencelium — architecture research, open under Apache 2.0 (code/weights) + CC BY 4.0 (paper/text)
- Running Laboratory of Absurd Ideas — small, odd hypotheses tested fast and reported honestly, positive or negative
- Sencelium — non-Transformer LM, self-monitoring signals, living memory pool
- Claunity — AI assistant inside the Unity Editor
- Neuropers — coming soon
