AI & ML interests
NeoMundi Research studies runtime signals for generative AI: stability, factual validity, semantic risk, cost and drift across models, tasks and professional contexts.
Recent Activity
NeoMundi Research
Runtime signals for generative AI systems.
NeoMundi Research develops and uses a runtime measurement instrument for generative AI systems. It instruments AI agents in real time, detects moments of drift, supports compliance workflows and produces recurring experimental cartographies
Our work focuses on:
- runtime stability
- factual validity
- semantic risk
- behavioral drift
- cost and efficiency
- professional use cases
- red team analysis
- instrument qualification
Research question
How can runtime signals help the AI ecosystem observe, compare and govern generative systems across real-world contexts?
Current focus
NeoMundi Research is building recurring experimental cartographies of generative AI systems.
The objective is not to produce a definitive ranking of AI models, but to qualify a runtime measurement instrument through open, inspectable and repeatable experiments.
Experimental program
Over the next 3 months, NeoMundi Research will publish experimental cohorts every 15 days.
Each cohort may include:
- anonymized model cartography
- professional segment testing
- RAG behavior analysis
- intra-provider comparisons
- red team cases
- contributor-led experiments
Principle
We do not classify models as globally good or bad.
We observe behavioral profiles across stability, validity, risk, cost and context.
Links
Website: https://neomundi.org
GitHub: https://github.com/neomundi-io
LLM Cartography: https://github.com/neomundi-io/llm-cartography