Mimir
A foundation model for dynamical systems. Given observed behavior, Mimir recovers the underlying mathematical structure — as readable, verifiable, composable programs.
We have foundation models for language, images, and code. We don't have one for the dynamical systems that describe the physical world — control systems, biological networks, signal processors, physical simulations.
Discovering the structure of these systems from observed behavior is one of the oldest problems in science. Existing approaches either fit black-box approximations (neural ODEs, system identification) or search symbolic spaces with hand-crafted heuristics (symbolic regression). Neither gives you what you actually want: a model that understands how dynamical systems compose, and can reason about structure the way a mathematician does.
Mimir is built on Gimle — a mathematical formalism that composes dynamical systems from a small set of primitive operators. Gimle provides the structured representation that makes the space learnable, paired with a simulator that can evaluate any expression exactly. This is the substrate Mimir learns over.
From Imitation to Reasoning
Mimir doesn't just pattern-match. It progresses through four stages of learning, each building a deeper capability — an approach inspired by how AlphaZero learned to play Go not by memorizing games, but by learning to reason about strategy.
Imitation
Next-token prediction on synthetic data. The model learns the language of compositional systems — what well-formed expressions look like and how structure maps to behavior.
Behavioral
A simulation-based loss replaces token-level supervision. The model learns to match what systems actually do — optimizing directly for functional correctness, not syntactic similarity.
Reinforcement
RL fine-tuning lets the model explore beyond the training distribution. It learns to search for better solutions through trial and reward, discovering compositions that supervised training never showed it.
Planning
Learned search over composition strategies. Rather than committing to a single prediction, the model evaluates candidate structures before selecting — learning how to think about composing systems, not just what to predict.
Fields of Application
Control Systems
Recover controller structure and feedback topologies from closed-loop behavior. Identify PID configurations, stability characteristics, and actuator-plant interactions from observed input-output data.
Climate & Weather
Discover the governing dynamics of atmospheric and ocean models from observational data. Recover interpretable compositional structure from complex, coupled geophysical systems.
Biology & Medicine
Infer regulatory networks, pharmacokinetics, and epidemiological compartment models from experimental measurements. Systems where interpretability matters as much as accuracy.
Finance
Model the compositional dynamics underlying market behavior. Discover feedback structures, mean-reversion mechanisms, and regime-switching patterns as interpretable, verifiable programs.
Physics & Engineering
Recover governing equations from sensor data or simulation output. Mechanical systems, fluid dynamics, thermodynamics — anywhere the laws are compositional but the structure is unknown.
Scientific Discovery
Automated hypothesis generation for dynamical systems. Let the model propose interpretable mathematical structures from data — structures that can be read, tested, and formally verified.