Searcharxiv⌕ Search

arXiv subjects

Hangyu Zhou

Publications and source records attributed to Hangyu Zhou.

9 recordsLinked to original sources
arXiv2026

The Garden of Forking Paths: Narrative Arc-Conditioned Gameplay Planning

Narrative archetypes (e.g., Hero's Journey, Three-act structure) provide universal story structures that resonate across cultures and media and are important for video game storytelling, yet existing LLM-based methods lack explicit use of these archetypes in procedurally generated games. We propose Forking Garden, a framework for narrative arc-conditioned gameplay planning that generates branching games from user-provided storylines. Our approach first generates a diverse pool of independent nodes, then assembles them into a dungeon graph via arc-guided constraint algorithms, where each node achieves multimodal alignment of gameplay elements. We develop an end-to-end interactive system that instantiates the framework.

cs.HC↗
arXiv2025

All Stories Are One Story: Emotional Arc Guided Procedural Game Level Generation

The emotional arc is a universal narrative structure underlying stories across cultures and media -- an idea central to structuralist narratology, often encapsulated in the phrase "all stories are one story." We present a framework for procedural game narrative generation that incorporates emotional arcs as a structural backbone for both story progression and gameplay dynamics. Leveraging established narratological theories and large-scale empirical analyses, we focus on two core emotional patterns -- Rise and Fall -- to guide the generation of branching story graphs. Each story node is automatically populated with characters, items, and gameplay-relevant attributes (e.g., health, attack), with difficulty adjusted according to the emotional trajectory. Implemented in a prototype action role-playing game (ARPG), our system demonstrates how emotional arcs can be operationalized using large language models (LLMs) and adaptive entity generation. Evaluation through player ratings, interviews, and sentiment analysis shows that emotional arc integration significantly enhances engagement, narrative coherence, and emotional impact. These results highlight the potential of emotionally structured procedural generation for advancing interactive storytelling for games.

cs.AI↗
arXiv2025

RanDeS: Randomized Delta Superposition for Multi-Model Compression

From a multi-model compression perspective, model merging enables memory-efficient serving of multiple models fine-tuned from the same base, but suffers from degraded performance due to interference among their task-specific parameter adjustments (i.e., deltas). In this paper, we reformulate model merging as a compress-and-retrieve scheme, revealing that the task interference arises from the summation of irrelevant deltas during model retrieval. To address this issue, we use random orthogonal transformations to decorrelate these vectors into self-cancellation. We show that this approach drastically reduces interference, improving performance across both vision and language tasks. Since these transformations are fully defined by random seeds, adding new models requires no extra memory. Further, their data- and model-agnostic nature enables easy addition or removal of models with minimal compute overhead, supporting efficient and flexible multi-model serving.

cs.LG↗
arXiv2025

Triple-Q state in magnetic breathing kagome lattice

Magnetic frustration in two-dimensional spin lattices with triangular motifs underpins a series of exotic states, ranging from multi-Q configurations to disordered spin-glasses. The antiferromagnetic kagome lattice, characterized by its network of corner-sharing triangles, represents a paradigmatic frustrated system exhibiting macroscopic degeneracy. Expanding upon the kagomerization mechanism, we focus on the magnetic breathing kagome lattice formed by a Mn monolayer deposited on a heavy metal substrate and capped with h-BN. The Mn kagome arrangement induces pronounced magnetic frustration, as evidenced by the nearly flat bands derived from spin spiral energy calculations. Including further-neighbor interactions reveals a spin spiral energy minimum along the $Γ$-K line and an intriguing triple-Q state with nonzero topological charge, potentially leading to highly nonlinear Hall effects. Furthermore, the flat band properties can further give rise to an even more complex spin configuration, marked by several Q-pockets in the spin structure factor. These results present a fertile ground for advancing the study of multi-Q states and exploring emergent topological phenomena.

cond-mat.mtrl-sci↗
arXiv2024

LIVE-GS: LLM Powers Interactive VR Experience with Physics-Aware Gaussian Splatting

As 3D Gaussian Splatting (3DGS) emerges as a leading approach for novel view synthesis and scene reconstruction, its potential in digital asset creation has gained significant attention. An increasing number of asset libraries based on GS are being established. However, generating physics-based dynamic assets remains a time-consuming and expertise-intensive task, especially for non-experts. In this paper, we propose LIVE-GS, a highly realistic Virtual Reality (VR) system powered by Large Language Models (LLMs), which enables rapid creation of dynamic Gaussian assets and real-time VR interactions. To inform our system design, we conducted interviews to examine challenges faced by current GS-based VR systems and the specific demands of users. Based on these insights, we employed GPT-4o to analyze key physical properties of objects that significantly impact user interactions, ensuring physics-based interactions in VR align with real-world phenomena. A key innovation of LIVE-GS is its ability to predict reasonable parameters in just 10 seconds from static Gaussian assets while maintaining high-quality VR interactions. To validate our approach, we invited participants experienced in physical simulation to manually adjust physical parameters, providing a baseline for comparison in both asset quality and authoring efficiency. We also conducted a comprehensive user study to evaluate system usability and user satisfaction. Experimental results demonstrate that LIVE-GS, leveraging LLMs' scene understanding capabilities, can achieve efficient physical scene creation and natural interactions without requiring manual design or annotation.

cs.HC↗
arXiv2024

AllClear: A Comprehensive Dataset and Benchmark for Cloud Removal in Satellite Imagery

Clouds in satellite imagery pose a significant challenge for downstream applications. A major challenge in current cloud removal research is the absence of a comprehensive benchmark and a sufficiently large and diverse training dataset. To address this problem, we introduce the largest public dataset -- $\textit{AllClear}$ for cloud removal, featuring 23,742 globally distributed regions of interest (ROIs) with diverse land-use patterns, comprising 4 million images in total. Each ROI includes complete temporal captures from the year 2022, with (1) multi-spectral optical imagery from Sentinel-2 and Landsat 8/9, (2) synthetic aperture radar (SAR) imagery from Sentinel-1, and (3) auxiliary remote sensing products such as cloud masks and land cover maps. We validate the effectiveness of our dataset by benchmarking performance, demonstrating the scaling law -- the PSNR rises from $28.47$ to $33.87$ with $30\times$ more data, and conducting ablation studies on the temporal length and the importance of individual modalities. This dataset aims to provide comprehensive coverage of the Earth's surface and promote better cloud removal results.

cs.CV↗
arXiv2023

Kagomerization of transition metal monolayers induced by two-dimensional hexagonal boron nitride

The kagome lattice is an exciting solid state physics platform for the emergence of nontrivial quantum states driven by electronic correlations: topological effects, unconventional superconductivity, charge and spin density waves, and unusual magnetic states such as quantum spin liquids. While kagome lattices have been realized in complex multi-atomic bulk compounds, here we demonstrate from first-principles a process that we dub kagomerization, in which we fabricate a two-dimensional kagome lattice in monolayers of transition metals utilizing a hexagonal boron nitride (h-BN) overlayer. Surprisingly, h-BN induces a large rearrangement of the transition metal atoms supported on a fcc(111) heavy-metal surface. This reconstruction is found to be rather generic for this type of heterostructures and has a profound impact on the underlying magnetic properties, ultimately stabilizing various topological magnetic solitons such as skyrmions and bimerons. Our findings call for a reconsideration of h-BN as merely a passive capping layer, showing its potential for not only reconstructing the atomic structure of the underlying material, e.g. through the kagomerization of magnetic films, but also enabling electronic and magnetic phases that are highly sought for the next generation of device technologies.

cond-mat.mtrl-sci↗
arXiv2022

Impact of Spin-Orbit Coupling on Quantum Transport in Magnetic Tunnel Junction with an anti-ferromagnetic Capping Layer

Using first-principles calculations, we explore the role of an anti-ferromagnetic heavy-metal, L1$_0$-IrMn, as a capping layer in a perpendicular magnetic tunnel junction (\emph{p}-MTJ). A comparative study is conducted by employing conventional non-magnetic heavy-metals (Ta, W or Mo) capping layers along with an anti-ferromagnetic IrMn in a symmetric-MTJ X/FeCo/MgO/FeCo/X, where X=Ta, W, Mo or IrMn. Firstly, the calculations without including spin-orbit coupling (SOC) are presented where the highest TMR is achieved in IrMn-IrMn MTJ compared to that of Ta-Ta, W-W and Mo-Mo MTJs. The origin of this large TMR is attributed to, both, the large spin-polarization due to reduced lattice-mismatch and the non-identical signatures of both the anti-parallel conduction-channels caused by the anti-ferromagnetic ordering of IrMn that reflects the spins at IrMn/FeCo interface. Moreover, when SOC is switched-on, the increase of TMR is observed in all the MTJs with a particularly giant enhancement in IrMn-IrMn MTJ. This SOC-induced increase in TMR is ascribed to the mixed contribution of $Δ_1$ and $Δ_5$ states and the additional increase in the band levels of the out-of-plane orbitals, \emph{p}$_z$, \emph{d}$_{z}^2$ and \emph{d}$_{xz}$ due to the lifting of degeneracy. Furthermore, it was also observed that the lattice-mismatch-induced strain might create an orbital reconstruction within the capping layer, probably, due to the crystallographic deformation and, in turn, in the adjacent FeCo-electrode. This microscopic mechanism also plays an additional decisive role in enhancing the TMR. Finally, our results indicate that IrMn can offer giant TMR in future spin-orbit toque (SOT) based MRAM devices with a straightforward design strategy.

cond-mat.mes-hall↗