Search Results for "learning"

Brax is a fast and fully differentiable physics engine for large-scale rigid body simulations, built on JAX. It is designed for research in reinforcement learning and robotics, enabling efficient simulations and gradient-based optimization.

MuJoCo, developed and maintained by Google DeepMind, is a high-performance physics engine designed for simulating complex, articulated systems that interact through contact. It is widely used in research fields such as robotics, biomechanics, computer graphics, animation, and machine learning, where fast and accurate physics simulations are essential. The engine provides a robust C API optimized for real-time computation, making it suitable for scientific research and advanced simulation environments. MuJoCo’s core architecture is performance-tuned and utilizes preallocated data structures created through an XML-based compiler. ...

Physical Symbolic Optimization (Φ-SO) - A symbolic optimization package built for physics. Symbolic regression module uses deep reinforcement learning to infer analytical physical laws that fit data points, searching in the space of functional forms.

A Java toolkit for the creation of physics simulations and 3D visualizations. Developed as part of the Technology Enabled Active Learning (TEAL) initiative at MIT.

Virtual interactive experiments (vexp) is a framework to promote the developmentand use of cooperative e-learning experiments over the internet. Interactive scientific experiments which can be controlled by various users synchronously.

An optical ray tracing program focused on layout of prototype instrumentation and learning about optics.

Jpowder, http://www.jpowder.org, is a Java viewer of powder diffraction data, and is a collaborative project between the ISIS Facility of the STFC and the E-Learning and Software Develepment Unit of the Kasem Bundit University.