arXiv subjects

Yunfei Gao

Publications and source records attributed to Yunfei Gao.

6 recordsLinked to original sources

arXiv2026

A Dual-Gate Altermagnetic Tunnel Junction Based on Bilayer Cr$_{2}$SeO

Altermagnets demonstrate significant potential in spintronics due to their unique non-relativistic spin-splitting properties, yet altermagnetic devices still face challenges in efficiently switching logic states. Here, we report electrostatically controllable spin-momentum locking in bilayer Cr$_{2}$SeO and design a dual-gate altermagnetic tunnel junction (AMTJ), which can switch between high and low resistance states without switching the Néel vector. First-principles calculations demonstrate that vertical electric field can induce significant spin splitting in bilayer Cr$_{2}$SeO. Reversing the electric field direction can alter the spin-momentum locking in bilayer Cr$_{2}$SeO. Leveraging this electric-field-tunable spin splitting, the dual-gate AMTJ exhibits an ultrahigh tunneling magnetoresistance (TMR) ratio of $10^{7}$. This work provides theoretical support for the design of fully electrically controlled AMTJs and demonstrates their great potential for applications in spintronic devices.

cond-mat.mtrl-sci↗

arXiv2025

From Diagnosis to Improvement: Probing Spatio-Physical Reasoning in Vision Language Models

Spatio-physical reasoning, a foundation capability for understanding the real physics world, is a critical step towards building robust world models. While recent vision language models (VLMs) have shown remarkable progress in specialized domains like multimodal mathematics and pure spatial understanding, their capability for spatio-physical reasoning remains largely unexplored. This paper provides a comprehensive diagnostic analysis of mainstream VLMs, revealing that current models perform inadequately on this crucial task. Further detailed analysis shows that this underperformance is largely attributable to biases caused by human-like prior and a lack of deep reasoning. To address these challenges, we apply supervised fine-tuning followed by rule-based reinforcement learning to Qwen2.5-VL-7B, resulting in significant improvements in spatio-physical reasoning capabilities and surpassing leading proprietary models. Nevertheless, despite this success, the model's generalization to new physics scenarios remains limited -- underscoring the pressing need for new approaches in spatio-physical reasoning.

cs.CV↗

arXiv2025

A Comprehensive Review on Noise Control of Diffusion Model

Diffusion models have recently emerged as powerful generative frameworks for producing high-quality images. A pivotal component of these models is the noise schedule, which governs the rate of noise injection during the diffusion process. Since the noise schedule substantially influences sampling quality and training quality, understanding its design and implications is crucial. In this discussion, various noise schedules are examined, and their distinguishing features and performance characteristics are highlighted.

cs.LG↗

arXiv2022

WILD-SCAV: Benchmarking FPS Gaming AI on Unity3D-based Environments

Recent advances in deep reinforcement learning (RL) have demonstrated complex decision-making capabilities in simulation environments such as Arcade Learning Environment, MuJoCo, and ViZDoom. However, they are hardly extensible to more complicated problems, mainly due to the lack of complexity and variations in the environments they are trained and tested on. Furthermore, they are not extensible to an open-world environment to facilitate long-term exploration research. To learn realistic task-solving capabilities, we need to develop an environment with greater diversity and complexity. We developed WILD-SCAV, a powerful and extensible environment based on a 3D open-world FPS (First-Person Shooter) game to bridge the gap. It provides realistic 3D environments of variable complexity, various tasks, and multiple modes of interaction, where agents can learn to perceive 3D environments, navigate and plan, compete and cooperate in a human-like manner. WILD-SCAV also supports different complexities, such as configurable maps with different terrains, building structures and distributions, and multi-agent settings with cooperative and competitive tasks. The experimental results on configurable complexity, multi-tasking, and multi-agent scenarios demonstrate the effectiveness of WILD-SCAV in benchmarking various RL algorithms, as well as it is potential to give rise to intelligent agents with generalized task-solving abilities. The link to our open-sourced code can be found here https://github.com/inspirai/wilderness-scavenger.

cs.LG↗

arXiv2010

Simulation of the Spin Field Effect Transistors: Effects of Tunneling and Spin Relaxation on its Performance

A numerical simulation of spin-dependent quantum transport for a spin field effect transistor (spinFET) is implemented in a widely used simulator nanoMOS. This method includes the effect of both spin relaxation in the channel and the tunneling barrier between the source/drain and the channel. Account for these factors permits setting more realistic performance limits for the transistor, especially the magnetoresistance, which is found to be lower compared to earlier predictions. The interplay between tunneling and spin relaxation is elucidated by numerical simulation. Insertion of the tunneling barrier leads to an increased magnetoresistance. Numerical simulations are used to explore the tunneling barrier design issues.

cond-mat.mes-hall↗

arXiv2010

p-i-n Tunnel FETs vs. n-i-n MOSFETs: Performance Comparison from Devices to Circuits

The band-to-band tunneling transistors have some performance advantages over the conventional MOSFETs due to the <60mV/dec sub-threshold slope. In this paper, carbon nanotubes are used as a model channel material to address issues that we believe will apply to BTBT FETs vs. MOSFETs more generally. We use pz-orbital tight-binding Hamiltonian and the non-equilibrium Green function (NEGF) formalism for rigorous treatment of dissipative quantum transport. A device level comparison of p-i-n TFETs and n-i-n MOSFETs in both ballistic and dissipative cases has been performed previously. In this paper, the possibility of using p-i-n TFETs in ultra-low power sub-threshold logic circuits is investigated using a rigorous numerical simulator. The results show that, in sub-threshold circuit operation, the p-i-n TFETs have better DC characteristics, and can deliver ~15x higher performance at the iso-P_LEAKAGE, iso-VDD conditions. Because p-i-n TFETs can operate at lower VDD than n-i-n MOSFETs, they can deliver ~3x higher performance at the same power (P_OPERATION). This results in ~3x energy reduction under iso-delay conditions. Therefore the p-i-n TFETs are more suitable for sub-threshold logic operation.

cond-mat.mes-hall↗