Applied AI Research Lab

Reasoning first. Learning second.

Centenum Labs is the research arm of Centenum Technologies. We work at the intersection of neurosymbolic computation, program synthesis, and causal modelling. Building AI that can be audited, corrected, and trusted at the level the decisions demand.

Read the Thesis Explore Research
Neurosymbolic Computation Program Synthesis Causal Modelling Symbolic Regression Structured Reasoning Auditable AI Reasoning First Neurosymbolic Computation Program Synthesis Causal Modelling Symbolic Regression Structured Reasoning Auditable AI Reasoning First
Research Areas
01
Neurosymbolic Computation

Combining learned representations with explicit, inspectable rules. Systems that can be audited, corrected, and trusted because their reasoning is inspectable by construction, not by post-hoc explanation.

Hybrid Models Interpretability Auditability
02
Program Synthesis

Turning intent into executable logic rather than probabilistic completion. Machines that construct verifiable programs from specifications, examples, and structured intent.

Executable Reasoning Verification Synthesis
03
Causal Modelling

Reasoning about what causes what, not just what correlates with what. Building systems that represent interventional structure so their predictions survive changes in the world they were trained on.

Causal Inference Interventions DAGs
04
𝕊
Symbolic Regression & Equation Discovery

Learning mathematical expressions directly from data. Symbolic regression as a bridge between raw ML and interpretable, human-readable models, especially in scientific and high-stakes domains.

PySR Equation Learning Physics-ML

AI that can
show its work.

04
Core research areas
01
Founding thesis
2025
Founded
LA·TO
Lagos · Toronto

The dominant bet in AI today is that optimising for likelihood is enough. Train a system to predict the next token across enough data, at enough scale, and a system you can trust with consequential decisions will emerge.

We are building on the assumption that it will not.

Reliable, auditable, domain-specific AI will not be trained into existence. It will be engineered, at the intersection of neurosymbolic computation, program synthesis, and causal modelling. Our parent company, Centenum Technologies, has shipped production AI systems since 2018. That ground truth shapes everything we build in the lab.

This is the compressed version. The full argument is set out in our founding thesis, Likelihood Is Not Truth.

Papers & Work

All Output ↗
Likelihood Is Not Truth
Centenum Labs · Founding Thesis · July 2026
Read →
001
Symbolic Regression as Interpretable Machine Learning: A Practitioner's Guide for Applied Sciences
Centenum Labs · 2026 · Symbolic Regression, ML Interpretability
Upcoming
002
From Data to Equations: Building MathExec, A Platform for Mathematical Modeling at Scale
Centenum Labs · 2024 · Platform Research, Applied ML
Upcoming
003
Library Learning in Program Synthesis: Surveying DreamCoder and Beyond
Centenum Labs · 2025 · Program Synthesis, Abstraction
Upcoming
004
Neurosymbolic Architectures for African Language Understanding
Centenum Labs · 2025 · NLP, African Languages, Hybrid AI
Upcoming
Open Tools
Mathematical Modeling Platform
MathExec

A platform for data scientists and researchers to go from paper to prototype, running symbolic regression, mathematical modeling, and equation discovery in the browser.

mathexec.com
Centenum Technologies · Lagos + Toronto · est. 2018
Research arm · centenumlabs.com
Stay in the Loop

Research,
distilled.

Low-frequency. High-signal. Updates on what we're building, reading, and thinking.