A self-learning scientific agent for X-ray diffraction
Gan Jiang improves X-ray diffraction analysis by rewriting and validating its own reusable skills, without retraining the model.
The paper says the agent sits on four diffraction tools covering phase identification, multiphase decomposition and whole-pattern modelling. Its frozen, self-improved skills beat the original expert-written skills across FullProf, GSAS-II and PyWPEM. The reported tests include overlapping reflections, a five-phase ancient Egyptian cosmetic, an operating battery and a disordered oxide catalyst. On DeltaXRDbench, it leads the evaluated methods for single- and multiphase identification across simulated and experimental data. HF Daily Papers' note
The paper says the agent sits on four diffraction tools covering phase identification, multiphase decomposition and whole-pattern modelling. Its frozen, self-improved skills beat the original expert-written skills across FullProf, GSAS-II and PyWPEM. The reported tests include overlapping reflections, a five-phase ancient Egyptian cosmetic, an operating battery and a disordered oxide catalyst. On DeltaXRDbench, it leads the evaluated methods for single- and multiphase identification across simulated and experimental data. HF Daily Papers' note
score 4