arXiv subjects

Siqi Shi

Publications and source records attributed to Siqi Shi.

4 recordsLinked to original sources

arXiv2026

A Hierarchical Synergistic Deep Learning Framework Integrating Composition, Structure, and Ionic Transport for Solid-State Electrolyte Discovery

Inorganic solid-state electrolytes must combine high room-temperature ionic conductivity, a wide electrochemical window, excellent electronic insulation, and favorable mechanical compliance. Single models struggle to support reliable multi-objective screening across vast chemical spaces because of training-data distribution mismatch, cross-property dataset heterogeneity, and scarce kinetic transport data. To overcome these limitations, we develop a hierarchical synergistic deep-learning framework that sequentially coordinates efficiency, accuracy, and reliability through four complementary modules. The in-house-developed L-G-DCNN and a multi-fidelity implementation built on DenseGNN serve as compositional and structural experts for thermodynamic coarse screening and multi-property evaluation, respectively; MatterSim and system-specific DeePMD models provide transport pre-assessment and kinetic validation. Systematic benchmarks show that each module outperforms mainstream counterparts in its task, while retrospective validation establishes dual closed-loop verification of module-level accuracy and end-to-end workflow reliability. Applied to 30,364,908 Alex/ICSD-derived candidates, the framework identifies 97 high-performance candidates with room-temperature ionic conductivities of 0.109--59.0 mS/cm, including 94 halides, one borohydride, and two oxides. Consistency with independent experimental data confirms that 76 of the 94 halides fall within reported high-conductivity structural regions. Analysis reveals that Li$^{+}$ jump-network connectivity, rather than the number of geometric Li sites, is the core determinant of room-temperature ionic conductivity. Li-defect engineering effectively enhances oxide transport, whereas the inherent rigidity of the O$^{2-}$ framework suggests a potential upper limit on oxide electrolyte performance.

cond-mat.mtrl-sci↗

arXiv2026

CLASH: Collision Learning via Augmented Sim-to-real Hybridization to Bridge the Reality Gap

The sim-to-real gap, particularly in the inaccurate modeling of contact-rich dynamics like collisions, remains a primary obstacle to deploying robot policies trained in simulation. Conventional physics engines often trade accuracy for computational speed, leading to discrepancies that prevent direct policy transfer. To address this, we introduce Collision Learning via Augmented Sim-to-real Hybridization (CLASH), a data-efficient framework that learns a parameter-conditioned impulsive collision surrogate model and integrates it as a plug-in module within a standard simulator. CLASH first distills a base model from an imperfect simulator (MuJoCo) using large-scale simulated collisions to capture reusable physical priors. Given only a handful of real collisions (e.g., 10 samples), it then (i) performs gradient-based identification of key contact parameters and (ii) applies small-step, early-stopped fine-tuning to correct residual sim-to-real mismatches while avoiding overfitting. The resulting hybrid simulator not only achieves higher post-impact prediction accuracy but also reduces the wall-clock time of collision-heavy CMA-ES search by 42-48% compared to MuJoCo. We demonstrate that policies obtained with our hybrid simulator transfer more robustly to the real world, doubling the success rate in sequential pushing tasks with reinforcement learning and significantly increase the task performance with model-based control.

cs.RO↗

arXiv2012

Microscopic origin of the structural phase transitions at the Cr2O3 (0001) surface

The surface of a Cr2O3 (0001) film epitaxially grown on Cr undergoes an unusual reentrant sequence of structural phase transitions. In order to understand the underlying microscopic mechanisms, the structural and magnetic properties of the Cr2O3 (0001) surface are here studied using first-principles electronic structure calculations. Two competing surface Cr sites are identified. The energetics of the surface is described by a configurational Hamiltonian with parameters determined using total energy calculations for several surface supercells. Effects of epitaxial strain and magnetic ordering on configurational interaction are also included. The thermodynamics of the system is studied using Monte Carlo simulations. At zero strain the surface undergoes an ordering phase transition at 165 K. Tensile epitaxial strain together with antiferromagnetic ordering drive the system toward strong configurational frustration, suggesting the mechanism for the disordering phase transition at lower temperatures.

cond-mat.mtrl-sci↗

arXiv2008

Magnetism of chromia

The electronic structure and magnetism of chromia (corundum-type Cr2O3) are studied using full-potential first-principles calculations. The electronic correlations are included within the LSDA+U method. The energies of different magnetic configurations are very well fitted by the Heisenberg Hamiltonian with strong exchange interaction with two nearest neighbors and additional weak interaction up to the fifth neighbor shell. These energies are insensitive to the position of the oxygen states, indicating that magnetism in Cr2O3 is dominated by direct exchange. The Neel temperature is calculated using the pair-cluster approximation for localized quantum spins of magnitude 3/2. Very good agreement with experiment is found for all properties including the equilibrium volume, spectral density, local magnetic moment, band gap, and the Neel temperature for the values of U and J that are close to those obtained within the constrained occupation method. The band gap is of the Mott-Hubbard type.

cond-mat.mtrl-sci↗