Ramblings
Diffusion Experiments for Generating Crystal Structures
Exploring hand-rolled diffusion models for crystal structure generation, focusing on larger supercells to enable more complex structures and error tolerance
https://ouro.foundation/posts/will/diffusion-experiments-for-generating-crystal-structures
Round-Trip Reinforcement Learning Experiments
Experimenting with round-trip reinforcement learning as a means to
help an LLM build a bidirectionally coherent internal model of
inorganic crystals
https://ouro.foundation/posts/will/round-trip-reinforcement-learning-experiments
Building a Physically Comparable Magnetic Hysteresis Simulation Framework
Overview of the current work and future enhancements for magnetic hysteresis simulations
https://ouro.foundation/posts/will/building-a-physically-comparable-magnetic-hysteresis-simulation-framework
Using MAE to Filter Permanent Magnet Candidates from Soft Magnets
Explanations on how MAE factors into a crucial permanent magnet property, coercivity, and how we can use calculated MAE values to get a good feel for which candidates have permanent magnet potential
https://ouro.foundation/posts/will/using-mae-to-filter-permanent-magnet-candidates-from-soft-magnets
Tree Searching Conclusions 1st Iteration
Diving into the results and limitations of the first tree search build. The generative step is a growing bottleneck
https://ouro.foundation/posts/will/tree-searching-conclusions-1st-iteration
Tree Searching for New Magnets
Adapting Sakana AI's work on tree search based prompt optimizers to intelligently navigate towards promising rare-earth free permanent magnet candidates
https://ouro.foundation/posts/will/tree-searching-for-new-magnets