Causal World Models

Physical AI needs causal foundations.

We build causal world models for physical simulation and robot learning — agents that reason about what an action will cause before taking it, instead of imitating demonstrations. The engine is live at app.decivine.com.

1 core

one causal engine,
simulation and control

2–3×

fewer training samples
to reach target skill

4 / 4

classic control
benchmarks solved

Action Outcome What If? Causal Agent Risk Policy Decision Consequence Environment World Model CF World Enrich Live Signal Real Data
What Ships Today

Live now at app.decivine.com

The platform builds a causal world model from your data, then predicts, answers counterfactuals, and lets a population of causal agents tune the model automatically. This is the engine we are now pointing at simulation and robot data.

Causal World Model from Your Data

Structured data or documents in — the platform extracts causal structure and builds a queryable graph. Intervene on any node and read off the downstream consequence.

Prediction & Counterfactual QA

Ask “what happens next?” or “what would happen if X changed?” — answered in one forward pass with the full causal chain in plain sight, restricted to verified causal edges rather than correlation.

Auto-Tuning Causal Agents

A population of causal agents trains on the world model, probes it where it is least certain, and writes what it learns back in — accuracy compounds with every pass.

Automated Enrichment

Once your graph exists, the platform extends it automatically — pulling in related entities, relationships and supporting evidence, so the model grows without manual curation.

Demo

See it in action

app.decivine.com

Bring your data. We’ll run a real counterfactual.

We walk design partners through the live platform on their own simulation or robot data — building the causal graph and running interventions in the session, not a scripted demo.

Book a live walkthrough →

Prefer to explore first? The platform is open at app.decivine.com.

Where We’re Going

The same core, aimed at where consequence is physical

Two markets, one engine. Robot learning is the wedge — every action is an intervention and every real trial is slow and costly, which is exactly where causal beats imitation. Physical simulation is the same buyer, out of the same budget line.

🤖

Robotics & Control

Agents that imagine an action’s outcome before committing — planning with counterfactual rollouts inside the world model, and reaching the same skill from a fraction of the training data.

First design partners

Reinforcement Learning

Every action is an intervention. Causal agents discover the structure of their environment and reason over it — better credit assignment, less confounding, far greater sample efficiency.

Validated

Physical Simulation

A causal world model of a physical system — contacts, friction, fluids, structural mechanics. Replace an expensive solver sweep with a single forward-pass query: “what if this parameter changed?”

On the roadmap
Early Proof

Live in production — and validated in research

We would rather be precise than impressive. Here is exactly what exists today, and exactly what it does not yet cover.

Live app

Causal world model, prediction, counterfactual QA and auto-tuning causal agents — running now at app.decivine.com

4 / 4

Classic control benchmarks — CartPole, Pendulum, MountainCar, LunarLander — matched the baseline policy at ~2–3× fewer training samples

Single pass

Counterfactual queries answered in one forward pass — no chain-of-thought, fast enough to sit inside a live control loop

Those four are simulated benchmarks, not physical hardware, and the result is a research finding separate from the live platform. Next milestone: contact-rich manipulation on real hardware with a design partner.

Get Involved

Back the causal world model

Early stage. Core validated, platform live. We’re signing the first design partners in robot learning and physical simulation — bring your data and we’ll run a real counterfactual, not a demo.

DECIVINE

Causal world models and the agentic systems that enrich them — one engine for physical simulation and robot learning.

© 2026 Decivine. All rights reserved.Causal · Counterfactual · Compounding