I do self-driven AI research and test small language architectures. Fan of hybrid models and gating
- AI Alignment & Transparency
- Value-Aligned AI
- (Hybrid)Language Model Architectures (SSMs; Hopfield; Latent Transformers)
- Time-Series Modeling of Psychological Behavior
- Autoregressive Time-Series Simulation Modeling of Astronaut Stress data, incl. Temporal Fusion Transformers
- 🎸 GearChat - RAG-powered chatbot for musicians to chat with their intruments manual (synthesizers, drum-machines etc.) (deployed)
-
🌌 CALM-Seal - A hybrid implementation of Shao et al. their Continuous Autoregressive Language Modeling (CALM) fused with State Space Models (SSM) and Hopfield Networks and gating + rectified flow.
-
🧠 Hamba - Novel hybrid hopfield + mamba inspired architecture combining SSMs and Hopfield networks with attention (BPE PPL of 16,2 (FineWiki) two hierarchical reasoning passes (Segmented-Reasoning) and rectified flow.
-
🏛️ MSC: RLAIF/RLHF for Public Value Alignment Enhancing Transparency in LLMs - Master's thesis on aligning Mistral-7b with democratic transparency values (VU Amsterdam, 2025)
-
📊 Time-Varying VAR/GAM and Temporal Fusion Models - Extension of my BSc extrapolated to astronauts on a 1 year mission. Predicting psychological behavior from simulated Astronaut data over a year with 3% shock event data (positive & negative) (VU University Amsterdam, 2024)
-
More can be seen in Repo's :-)
- 📊 BSC: Time-Varying Non-linear Emotion Modelling: Vector Autoregression and Generalized Additive Models Models - Predicting psychological behavior from EMA data (Soyster et al., 2020) (Maastricht University, 2022)
ML/AI: PyTorch, JAX/Flax, Transformers, TRL, LangChain , FAISS, ChromaDB, Mamba, Energy-Transformers, CALM, Optuna Data Science: R, pandas, statsmodels, mgm, Time-Series, Markov Chain, Time-Varying VAR, Time-Varying Generalized Additive Models Deployment: Cloud GPUs, REST APIs, vector databases Random fact curious about mojo.
[](will add later)
"Technical sophistication alone cannot resolve fundamental tensions in democratic governance"
— From my thesis on AI transparency