Computational materials scientist · Postdoctoral researcher · Ph.D. Jilin University (2025)
First-principles calculations, machine-learning potentials, and LLM-agent systems for data-driven materials design: halide perovskite interfaces, self-trapped excitons in metal halides, and interpretable descriptors.
My research code lives in the Material Intelligence organization, which I founded and lead.
- elf-anisotropy-sisso — ELF anisotropy descriptor CLI and SISSO++ workflow for 2D iodide perovskites
- derivationlab — LLM-agent physics derivations with verifiable, hash-chained records
- materials-discovery-skills — generative design, MLIP prescreening and DFT validation of inorganic semiconductors, packaged as agent skills
- perovskite-adsorption-sampler — adsorbate placement and MLP-to-DFT screening on perovskite surfaces
- mcp-zotero — my fork of the Zotero MCP server (group-library support)