Get Started
Home
Topics
Search
Library
Research questionHow can differentiable, structure-preserving finite-time maps avoid secular phase drift in long-duration dissipative inspirals?Fast orbital motion and slow dissipative evolution occur on different timescales, so small errors in each finite-time update can accumulate into secular phase drift over many compositions.
AI
Machine Learning
Neural and Evolutionary Computing
Research Paper
Technology
Latest papersRecent research connected to this question, newest first.Introducing SINFONIA: Symplectic, slimplectic and Magnusian (Neural) Flows for Orbital Numerical Integration and AccelerationThe evidence is a proof of concept for a 2.5PN neutron-star inspiral. It evaluates symplectic/slimplectic, Taylor-anchored, and Magnusian neural-flow architectures over 10^2–10^5 map compositions at timesteps of a full orbital period and beyond, with an additional force-inference experiment; broader orbital systems are not established by the source.research paper · Sep 3, 2026
Related questions
How can we simulate and backpropagate through long coupled dynamical systems without stepwise time recursion?How can learned chaotic systems preserve long-term statistics and stability beyond short-term trajectory matching?How can unknown nonlinear ODE dynamics and time-varying external forces be learned from one trajectory without physics priors?How can learned graph simulators preserve contact history for accurate long-horizon granular-dynamics rollouts?