arXiv subjects

Ruichao Chen

Publications and source records attributed to Ruichao Chen.

2 recordsLinked to original sources

arXiv2026

Superconductivity in strongly overdoped cuprates: beyond the single-band model

In order to explain the observation of an extended superconducting region in several overdoped cuprates, which contrasts the dome scenario, by means of neutron and synchrotron x-ray powder diffraction we study the crystal structure of YBa$_2$Cu$_3$O$_{y}$, where strong oxygen overdoping up to $y = 7.4$ is achieved under high-pressure. A bond valence sum analysis indicates that 1/5 of the extra holes created by the excess oxygen are transferred to the CuO$_2$ planes, thus increasing the hole density up to $p=0.27$ hole/Cu, where superconductivity is expected to vanish according to the dome scenario. Instead, our data confirm a previous observation [Okai, Ono and Mitsuhashi, Physica C: Superconductivity {\bf 366}, 164 (2002)] that the superconducting critical temperature, $T_c$, remains constant with $y$. Our data analysis accounts for this discrepancy in terms of the much shorter bond between the apical oxygen and the planar Cu ion, which suggests that the extra holes occupy the $a_1$-symmetry states formed by $d_{3z^2-r^2}$ orbitals, instead of the usual $b_1$-symmetry Zhang-Rice singlet states formed by $d_{x^2-y^2}$ orbitals. Suitable spectroscopic measurements on single crystals may support such a two-band scenario, which would require a totally different theoretical approach to explain superconductivity in cuprates.

cond-mat.supr-con↗

arXiv2026

Med-Scout: Curing MLLMs' Geometric Blindness in Medical Perception via Geometry-Aware RL Post-Training

Despite recent Multimodal Large Language Models (MLLMs)' linguistic prowess in medical diagnosis, we find even state-of-the-art MLLMs suffer from a critical perceptual deficit: geometric blindness. This failure to ground outputs in objective geometric constraints leads to plausible yet factually incorrect hallucinations, rooted in training paradigms that prioritize linguistic fluency over geometric fidelity. This paper introduces Med-Scout, a novel framework that "cures" this blindness via Reinforcement Learning (RL) that leverages the intrinsic geometric logic latent within unlabeled medical images. Instead of relying on costly expert annotations, Med-Scout derives verifiable supervision signals through three strategic proxy tasks inspired by the systematic reading and reasoning patterns of clinicians: Hierarchical Scale Localization, Topological Jigsaw Reconstruction, and Anomaly Consistency Detection. To rigorously quantify this deficit, we present Med-Scout-Bench, a new benchmark specifically designed to evaluate geometric perception. Extensive evaluations show that Med-Scout significantly mitigates geometric blindness, outperforming leading proprietary and open-source MLLMs by over 40% on our benchmark. Furthermore, this enhanced geometric perception generalizes to broader medical understanding, achieving superior results on radiological and comprehensive medical VQA tasks.

cs.CV↗