The ALR-Aerospace model series

10/08/2026By Alrithm Engineering
The ALR-Aerospace model series

The ALR-Aerospace model series

Alrithm Team

The aerospace platform is live. Fourteen categories, each with 200 verified rows, all served through one streaming endpoint.

What it covers

Computational categories — the answer is a number, and the number comes with a derivation.

CategoryWhat it teaches
Propulsion and Mass BudgetRocket equation, propellant mass, staging, finite-burn gravity losses
Orbital MechanicsCircular-orbit period, Hohmann transfer, Kepler's equation
Orbit PerturbationsFirst-order J2 nodal regression, sun-synchronous inclination
Launch Geometry and WindowsLaunch azimuth with and without Earth rotation, plane-crossing wait time, range azimuth margins
Standard AtmosphereUS Standard Atmosphere 1976 density and dynamic pressure to 20 km
HardwareIdeal nozzle exit velocity, thick-wall tank thickness
Guidance and EstimationKalman filter predict-and-update cycle
Ascent SimulationBurnout state and Max-Q, converged across three numerical integrators

Refusal categories — the answer is a code, and the code names what's missing.

CategoryThe model learns to
Refusals: Missing ParameterAsk for the specific field it needs
Refusals: UnderspecifiedAsk which interpretation the user meant
Refusals: Invalid InputAsk for a corrected value
Refusals: Inconsistent InputAsk the user to reconcile the contradiction
Refusals: InfeasibleSay the goal has no solution instead of inventing a number
Refusals: Legal and LicensingRoute the question to the correct authority

Why the refusal categories matter

A model trained only on solvable problems learns that every question has an answer. Given a question with a missing input, it invents one. Given contradictory constraints, it picks one silently. The refusal categories teach the opposite: stop, name what's wrong, ask.

That's the difference between a calculator that returns garbage and an assistant that says "I need the chamber pressure before I can compute exit velocity."

Verification

Every computational row carries:

  • headline_answer — the display value
  • headline_answer_exact — the same value at 40-digit precision
  • error_bound — the maximum deviation from the true answer
  • verification_level — analytical_proof, numerical_convergence, or interval_arithmetic
  • neglected_effects — what the model omits (J2, drag, Earth rotation)

A row is accepted only if the analytical half-width is at most 6.2×10⁻⁴, the delivered code agrees with the reference within 2.7×10⁻¹¹, and the independent numerical method agrees within 1.5×10⁻¹⁰.

Ascent simulation rows are verified differently. They run three independent integrators — RK4, DOP853, Radau — and accept the result only when all three agree within 1×10⁻⁷. The row's fidelity_tier field says which method was used.

How to stream it

GET /api/v1/models/alr-aerospace-platform/data

Same key, same format, same query parameters as the code platform. Filter by category with sample_type:

GET /api/v1/models/alr-aerospace-platform/data?sample_type=orbital_mechanics

What's next

More templates per category. Deeper coverage of launch windows, more staging configurations, and expanded perturbation cases.

— The Alrithm team

Associated Data Model

This article references the ALR Aerospace Platform 0.2. You can stream these verified reasoning traces immediately using our API.

[ FETCH DATA VIA API ]