I'm a Postdoctoral Research Associate at Northeastern University, working in the Rad Lab and Network Science Institute (NetSI). My research sits at the intersection of machine learning, network science, and human behavior.
- 🧠 Veracity in Large Language Models (LLMs): understanding how models encode factual inforamtion.
- 🌐 Networked Intelligence: modeling belief diffusion and collective reasoning in interconnected AI-human systems.
- ⚖️ Interpretability: building transparent, trustworthy ML tools for behavioral and social data.
- 💡 Modeling Trajectories with LLMs
- Trilemma of Truth: dataset and analysis framework exploring how LLMs represent and use factual information.
- sAwMIL (Sparse-Aware Multiple Instance Learning): Multiple-Instance SVM classifier
- LIFE2VEC and life2vec-light: transformer model for life-course prediction using population-scale registry data (developed during my PhD).
- Savcisens, Germans, et al. "Using sequences of life-events to predict human lives." Nature Computational Science 4.1 (2024): 43-56.
- Savcisens, Germans, and Tina Eliassi-Rad. "Trilemma of Truth in Large Language Models." Mechanistic Interpretability Workshop at NeurIPS 2025. 2025.
(Full list on Google Scholar or ORCID.)
- 🌍 savcisens.com
- 🧵 Bluesky
- 📚 Google Scholar
P.S. Carlo Marx is a character in Jack Kerouac's novel On the Road ...and DK? Honestly, I created this account in Denmark, and now it’s part of my lore.