Searcharxiv⌕ Search

arXiv subjects

Chenyan Wang

Publications and source records attributed to Chenyan Wang.

6 recordsLinked to original sources

Graph diffusion transformer for interpretable generation and inverse design of disordered atomic structures

Atomic disorder offers a potent degree of freedom to tailor material properties, yet the resulting vast configuration space obscures structure-property relations and hinders rational inverse design. Here we introduce Graph Amorphous Diffusion Transformer (GaDiT), a generative model that combines dynamic periodic graphs, masked conditioning and an interval-weighted training-noise distribution. GaDiT generates high-quality disordered structures while accelerating sampling by two orders of magnitude compared with existing corrector-based approaches, as demonstrated across two-dimensional amorphous carbon and three-dimensional amorphous silicon and silica. Moreover, GaDiT supports conditional generation under single or combined structural features, physical properties and preparation conditions in a unified model, and enables inverse design that expands accessible structure-property ranges beyond the training distribution. To interpret these structures and the trained network, we propose a conditioned atomic response mapping analysis, revealing both the global structural descriptors and the local atomic environments that govern target properties. Overall, GaDiT establishes a general framework for interpretable generation, inverse design and structure-property discovery for disordered atomic structures.

cond-mat.mtrl-sci↗

Beyond Benchmark Accuracy: A Diagnostic Test for LLM Chemical Prediction

LLMs have achieved strong performance on chemical prediction benchmarks. However, existing evaluation methods typically provide target entity identity information directly, making it difficult to determine whether model predictions arise from transferable structure-property reasoning capabilities or rely on target-related contextual information or memory retrieval. We propose the Identity-Blinded Reasoning Diagnostic framework to evaluate chemical prediction models when target entity identity information is hidden. In this framework, the target is no longer directly defined by its entity identity, but instead by a known source entity and an explicit structural edit operation. This setting preserves the information required to determine the target while reducing the opportunity for models to access target identity cues. We evaluate this framework across six tasks spanning odor prediction, molecular property prediction, retrosynthesis, and reaction yield prediction. The results show that models generally achieve better prediction performance when source and target entity properties are consistent. Providing explicit source entity labels further strengthens this dependence. Moreover, while keeping the target entity, structural change, and ground-truth label unchanged, replacing only the source entity label causes some model predictions to systematically shift toward the substituted source labels. These results indicate that high benchmark accuracy in chemical prediction tasks does not necessarily imply that LLMs possess reliable structure-property reasoning capabilities. Our diagnostic framework provides a controlled evaluation approach for testing whether models can genuinely update predictions according to entity changes. The framework can be extended to other AI for Science tasks, particularly those where targets can be defined through transformations of known entities.

cs.CL↗

MolHop: Scaffold-Hopping Molecular Design with Agentic Reinforcement Learning

Molecular design is often formulated as an optimization problem over existing molecular spaces. However, molecular optimization and the discovery of structurally distinct replacement molecules represent two fundamentally different search problems. LLM-based molecular design agents typically perform the former, conducting local refinement within the neighborhood of existing structures to improve target properties through structural modifications. In contrast, the latter requires exploration of a broader chemical space to discover molecules with entirely different scaffolds while preserving the functional characteristics of reference molecules. We propose MolHop, an agentic framework for cross-scaffold molecular discovery that enables models to navigate and edit within a pre-organized chemical space. MolHop organizes approximately 1.6 million molecular scaffolds from commercial molecules into a coarse-to-fine hierarchical action space. The agent selects new exploration regions through scaffold navigation actions and adjusts candidate properties through molecular editing. The framework maintains historical trajectory memory to support long-horizon search processes. On drug molecule and commercial compound tasks, connecting Grok-4.6 and DeepSeek-Flash-V4.1 to the MolHop framework substantially improves their cross-scaffold joint success rates. We develop MolHop-9B through agentic reinforcement learning, and MolHop-9B outperforms generative baselines including GPT-5.6-sol, Claude-5-Opus, and Gemini-3.1-Pro, while generating candidate molecules with larger scaffold differences from the reference molecules. These results demonstrate that effective molecular discovery beyond the traditional optimization paradigm requires structured chemical and action spaces, enabling agents to explore new structural regions rather than merely refine existing molecules.

cs.LG↗

Beyond overcomplication: a linear model suffices to decode hidden structure-property relationships in glasses

Establishing reliable and interpretable structure-property relationships in glasses is a longstanding challenge in condensed matter physics. While modern data-driven machine learning techniques have proven highly effective in establishing structure-property correlations, many models are criticized for lacking physical interpretability and being task-specific. In this work, we identify an approximate linear relation between structure profiles and disorder-induced responses of glass properties based on first order perturbation theory. We analytically demonstrate that this relationship holds universally across glassy systems with varying dimensions and distinct interaction types. This robust theoretical relationship motivates the adoption of linear machine learning models, which we show numerically to achieve surprisingly high predictive accuracy for structure-property mapping in a wide variety of glassy materials. We further devise regularization analysis to further enhance the interpretability of our model, bridging the gap between predictive performance and physical insight. Overall, this linear relation establishes a simple yet powerful connection between structural disorder and spectral properties in glasses, opening a new avenue for advancing their studies.

cond-mat.dis-nn↗

Integral Variable Range Hopping for Modeling Electrical Transport in Disordered Systems

The variable range hopping (VRH) model has been widely applied to describe electrical transport in disordered systems, providing theoretical formulas to fit temperature-dependent electric conductivity. These models rely on oversimplified assumptions that restrict their applicability and result in problematic fitting behaviors, yet their overusing situation is becoming increasingly serious. In this work we formulate an integral variable range hopping (IVRH) model, which replaces the empirical temperature power-law dependence in standard VRH theories with a physics-inspired integral formulation. The model builds upon the standard hopping probability $ω(R)$ w.r.t. hopping distance $R$ and incorporates the density of accessible electronic states through an effective volume function $V(R)$, which reflects the influence of system geometry. The IVRH formulation inherently reproduces both the Mott behavior at low temperatures and the Arrhenius behavior at high temperatures, respectively, and enables a smooth transition between the two regimes. We apply the IVRH model to two-dimensional, three-dimensional, and multi-layered systems. Monte Carlo simulations validate the model's predictions and yield consistent values for the fitting parameters, with substantially reduced variances compared to fitting using the standard VRH model. Furthermore, the improved robustness of IVRH also extends to the transport measurements in monolayer MoS$_2$ system and monolayer WS$_2$ system, enabling more physically meaningful interpretation.IVRH model offers a more stable and physically sound framework for interpreting hopping transport in low-dimensional amorphous materials, providing deeper insights into the universal geometric scaling factors that govern charge transport in disordered systems.

cond-mat.dis-nn↗

Predicting macroscopic properties of amorphous monolayer carbon via pair correlation function

Establishing the structure-property relationship in amorphous materials has been a long-term grand challenge due to the lack of a unified description of the degree of disorder. In this work, we develop SPRamNet, a neural network based machine-learning pipeline that effectively predicts structure-property relationship of amorphous material via global descriptors. Applying SPRamNet on the recently discovered amorphous monolayer carbon, we successfully predict the thermal and electronic properties. More importantly, we reveal that a short range of pair correlation function can readily encode sufficiently rich information of the structure of amorphous material. Utilizing powerful machine learning architectures, the encoded information can be decoded to reconstruct macroscopic properties involving many-body and long-range interactions. Establishing this hidden relationship offers a unified description of the degree of disorder and eliminates the heavy burden of measuring atomic structure, opening a new avenue in studying amorphous materials.

cond-mat.mtrl-sci↗