Composable by default
Metrics, datasets, and uncertainty estimates work together without requiring custom pipelines.
Software
Open-source tools for AI evaluation.
We build open-source PyTorch tools for psychometric modeling, adaptive testing, and rigorous evaluation of AI systems.
Measurement-aware outputs
Every output makes its assumptions visible. Results always include uncertainty and comparability information.
Built for community use
Researchers across institutions can adopt, extend, and contribute back to shared infrastructure.
Libraries
Library
torch_measure
A PyTorch library for measurement science. Includes IRT models (Rasch, 2PL, 3PL), computerized adaptive testing, psychometric metrics, and GPU-accelerated estimation.