arXiv subjects

Kai Gu

Publications and source records attributed to Kai Gu.

12 recordsLinked to original sources

arXiv2026

Unraveling the Size Determination Mechanism of Nanocrystal Synthesis via Interpretable Neural Networks

Deep learning models of nanocrystal synthesis enable the prediction of size and shape by encoding precursors and reaction conditions. However, their black-box nature hinders gaining deep insights into the underlying synthetic mechanisms. Here, we develop the Nanocrystal Equation Learner (NanoEQL), a fully white-box neural network to unravel the size determination mechanisms of nanocrystal synthesis. Building on the EQL architecture, eight operators are introduced to replace standard activation functions to fit the mathematical equations in nanocrystal synthesis. Among these operators, three smoothed operators address the gradient explosion of singular operators at zero. To evaluate the weights of different precursors, we develop a temperature-gated attention pooling strategy that encodes concentration-driven and reactivity-driven chemical synthesis mechanisms into the temperature gate. The NanoEQL model illustrates that the final nanocrystal size can be described by a linear equation composed of three scalars representing nanocrystallization capability (-Zp), growth capability (Zrea), and external input potential (-Zops). These interpretable scalars not only advance the rational design of nanocrystal synthesis but also establish a generalizable paradigm for deciphering chemical reaction mechanisms through white-box machine learning.

cs.LG↗

arXiv2026

Answering Without Referring: How AI Search Rewrites the Web's Economic Bargain

Search engines have long allocated attention on the web by routing users from queries to websites. AI search changes this arrangement because information needs can be resolved inside the intermediary. Using URL-level Comscore U.S. desktop clickstream, we compare ChatGPT and Google information-seeking occasions and exploit ChatGPT Search access expansions to estimate traditional search displacement. ChatGPT produces outbound clicks in only 5.2% of conversation sessions, far below Google's referral ratio. The remaining clicks are not a scaled-down Google stream: they skew toward specialized destinations and away from ad-supported sites. Wider access cuts search use by 9.4%, with search-referral losses largest for informational categories. Our findings identify a central economic shift in digital intermediation: AI search might satisfy information needs inside the intermediary while weakening the referral bargain that has linked search, traffic, and content production on the open web.

cs.CY↗

arXiv2026

A large-scale nanocrystal database with aligned synthesis and properties enabling generative inverse design

The synthesis of nanocrystals has been highly dependent on trial-and-error, due to the complex correlation between synthesis parameters and physicochemical properties. Although deep learning offers a potential methodology to achieve generative inverse design, it is still hindered by the scarcity of high-quality datasets that align nanocrystal synthesis routes with their properties. Here, we present the construction of a large-scale, aligned Nanocrystal Synthesis-Property (NSP) database and demonstrate its capability for generative inverse design. To extract structured synthesis routes and their corresponding product properties from literature, we develop NanoExtractor, a large language model (LLM) enhanced by well-designed augmentation strategies. NanoExtractor is validated against human experts, achieving a weighted average score of 88% on the test set, significantly outperforming chemistry-specialized (3%) and general-purpose LLMs (38%). The resulting NSP database contains nearly 160,000 aligned entries and serves as training data for our NanoDesigner, an LLM for inverse synthesis design. The generative capability of NanoDesigner is validated through the successful design of viable synthesis routes for both well-established PbSe nanocrystals and rarely reported MgF2 nanocrystals. Notably, the model recommends a counter-intuitive, non-stoichiometric precursor ratio (1:1) for MgF2 nanocrystals, which is experimentally confirmed as critical for suppressing byproducts. Our work bridges the gap between unstructured literature and data-driven synthesis, and also establishes a powerful human-AI collaborative paradigm for accelerating nanocrystal discovery.

cond-mat.mtrl-sci↗

arXiv2025

Negative Marginal Densities in Mixed Quantum-Classical Liouville Dynamics

The mixed quantum-classical Liouville equation (QCLE) provides an approximate perturbative framework for describing the dynamics of systems with coupled quantum and classical degrees of freedom of disparate thermal wavelengths. The evolution governed by the Liouville operator preserves many properties of full quantum dynamics, including the conservation of total population, energy, and purity, and has shown quantitative agreement with exact quantum results for the expectation values of many observables where direct comparisons are feasible. However, since the QCLE density matrix operator is obtained from the partial Wigner transform of the full quantum density matrix, its matrix elements can have negative values, implying that the diagonal matrix elements behave as pseudo-densities rather than densities of classical phase space. Here, we compare phase-space distributions generated by exact quantum dynamics with those produced by QCLE evolution from pure quantum initial states. We show that resonance effects in the off-diagonal matrix elements differ qualitatively, particularly for low-energy states. Furthermore, numerical and analytical results for low-dimensional models reveal that the QCLE can violate the positivity of marginal phase-space densities, a property that should hold at all times for any physical system. A perturbative analysis of a model system confirms that such violations arise generically. We also show that the violations of positivity of the marginal densities vanish as the initial energy of the system increases relative to the energy gap between subsystem states. These findings suggest that a negativity index, quantifying deviations from positivity, may provide a useful metric for assessing the validity of mixed quantum\textendash{}classical descriptions.

quant-ph↗

arXiv2025

Diagnosing Shortcut-Induced Rigidity in Continual Learning: The Einstellung Rigidity Index (ERI)

Deep neural networks frequently exploit shortcut features, defined as incidental correlations between inputs and labels without causal meaning. Shortcut features undermine robustness and reduce reliability under distribution shifts. In continual learning (CL), the consequences of shortcut exploitation can persist and intensify: weights inherited from earlier tasks bias representation reuse toward whatever features most easily satisfied prior labels, mirroring the cognitive Einstellung effect, a phenomenon where past habits block optimal solutions. Whereas catastrophic forgetting erodes past skills, shortcut-induced rigidity throttles the acquisition of new ones. We introduce the Einstellung Rigidity Index (ERI), a compact diagnostic that disentangles genuine transfer from cue-inflated performance using three interpretable facets: (i) Adaptation Delay (AD), (ii) Performance Deficit (PD), and (iii) Relative Suboptimal Feature Reliance (SFR_rel). On a two-phase CIFAR-100 CL benchmark with a deliberately spurious magenta patch in Phase 2, we evaluate Naive fine-tuning (SGD), online Elastic Weight Consolidation (EWC_on), Dark Experience Replay (DER++), Gradient Projection Memory (GPM), and Deep Generative Replay (DGR). Across these continual learning methods, we observe that CL methods reach accuracy thresholds earlier than a Scratch-T2 baseline (negative AD) but achieve slightly lower final accuracy on patched shortcut classes (positive PD). Masking the patch improves accuracy for CL methods while slightly reducing Scratch-T2, yielding negative SFR_rel. This pattern indicates the patch acted as a distractor for CL models in this setting rather than a helpful shortcut.

cs.LG↗

arXiv2025

Mano Technical Report

Graphical user interfaces (GUIs) are the primary medium for human-computer interaction, yet automating GUI interactions remains challenging due to the complexity of visual elements, dynamic environments, and the need for multi-step reasoning. Existing methods based on vision-language models (VLMs) often suffer from limited resolution, domain mismatch, and insufficient sequential decisionmaking capability. To address these issues, we propose Mano, a robust GUI agent built upon a multi-modal foundation model pre-trained on extensive web and computer system data. Our approach integrates a novel simulated environment for high-fidelity data generation, a three-stage training pipeline (supervised fine-tuning, offline reinforcement learning, and online reinforcement learning), and a verification module for error recovery. Mano demonstrates state-of-the-art performance on multiple GUI benchmarks, including Mind2Web and OSWorld, achieving significant improvements in success rate and operational accuracy. Our work provides new insights into the effective integration of reinforcement learning with VLMs for practical GUI agent deployment, highlighting the importance of domain-specific data, iterative training, and holistic reward design.

cs.MM↗

arXiv2025

Alcohol induced surface charging of colloidal quantum dots for controllable electrophoretic deposition processing

In this work, we report an alcohol-induced surface charging route of colloidal QDs to achieve controllable electrophoretic deposition processing. By adding a fixed amounts of alcohols into a preformed quantum dots solution in non-polar solvents, the colloidal quantum dots can be positively charged, and then deposited on negative electrode under applied electric field. The surface charging of PbSe quantum dots was investigated by zeta potential, nuclear magnetic resonance, Fourier transform infrared spectroscopy, and discrete Fourier transform calculations. It was found that the zeta potential of oleate acid capped PbSe QDs increases from +1.6 mV to +13.4 mV with the amount of alcohol solvent increasing. The alcohol-induced zeta potential increasing can be explained to the electron cloud shift of active hydrogen mediated by intermolecular hydrogen bonds between carboxy acid and alcohol. Considering the influence of surface charging of quantum dots on their dispersibility, we describe the microscopic mechanism of alcohol-induced electrophoretic deposition processing. Furthermore, we developed a size-selective separation protocol by controlling alcohol-induced electrophoretic deposition processing.

physics.app-ph↗

arXiv2024

Deep Learning Models for Colloidal Nanocrystal Synthesis

Colloidal synthesis of nanocrystals usually includes complex chemical reactions and multi-step crystallization processes. Despite the great success in the past 30 years, it remains challenging to clarify the correlations between synthetic parameters of chemical reaction and physical properties of nanocrystals. Here, we developed a deep learning-based nanocrystal synthesis model that correlates synthetic parameters with the final size and shape of target nanocrystals, using a dataset of 3500 recipes covering 348 distinct nanocrystal compositions. The size and shape labels were obtained from transmission electron microscope images using a segmentation model trained with a semi-supervised algorithm on a dataset comprising 1.2 million nanocrystals. By applying the reaction intermediate-based data augmentation method and elaborated descriptors, the synthesis model was able to predict nanocrystal's size with a mean absolute error of 1.39 nm, while reaching an 89% average accuracy for shape classification. The synthesis model shows knowledge transfer capabilities across different nanocrystals with inputs of new recipes. With that, the influence of chemicals on the final size of nanocrystals was further evaluated, revealing the importance order of nanocrystal composition, precursor or ligand, and solvent. Overall, the deep learning-based nanocrystal synthesis model offers a powerful tool to expedite the development of high-quality nanocrystals.

cond-mat.mtrl-sci↗

arXiv2024

RobKiNet: Robotic Kinematics Informed Neural Network for Optimal Robot Configuration Prediction

Task and Motion Planning (TAMP) is essential for robots to interact with the world and accomplish complex tasks. The TAMP problem involves a critical gap: exploring the robot's configuration parameters (such as chassis position and robotic arm joint angles) within continuous space to ensure that task-level global constraints are met while also enhancing the efficiency of subsequent motion planning. Existing methods still have significant room for improvement in terms of efficiency. Recognizing that robot kinematics is a key factor in motion planning, we propose a framework called the Robotic Kinematics Informed Neural Network (RobKiNet) as a bridge between task and motion layers. RobKiNet integrates kinematic knowledge into neural networks to train models capable of efficient configuration prediction. We designed a Chassis Motion Predictor(CMP) and a Full Motion Predictor(FMP) using RobKiNet, which employed two entirely different sets of forward and inverse kinematics constraints to achieve loosely coupled control and whole-body control, respectively. Experiments demonstrate that CMP and FMP can predict configuration parameters with 96.67% and 98% accuracy, respectively. That means that the corresponding motion planning can achieve a speedup of 24.24x and 153x compared to random sampling. Furthermore, RobKiNet demonstrates remarkable data efficiency. CMP only requires 1/71 and FMP only requires 1/15052 of the training data for the same prediction accuracy compared to other deep learning methods. These results demonstrate the great potential of RoboKiNet in robot applications.

cs.RO↗

arXiv2023

Error Analysis of Mixed Residual Methods for Elliptic Equations

We present a rigorous theoretical analysis of the convergence rate of the deep mixed residual method (MIM) when applied to a linear elliptic equation with various types of boundary conditions. The MIM method has been proposed as a more effective numerical approximation method compared to the deep Galerkin method (DGM) and deep Ritz method (DRM) in various cases. Our analysis shows that MIM outperforms DRM and deep Galerkin method for weak solution (DGMW) in the Dirichlet case due to its ability to enforce the boundary condition. However, for the Neumann and Robin cases, MIM demonstrates similar performance to the other methods. Our results provides valuable insights into the strengths of MIM and its comparative performance in solving linear elliptic equations with different boundary conditions.

math.NA↗

arXiv2019

Cooperative Vision-based Localization Networks with Communication Constraints

Accurate location information is indispensable for the emerging applications of \ac{iov}, such as automatic driving and formation control. In the real scenario, vision-based localization has demonstrated superior performance to other localization methods for its stability and flexibility. In this paper, a scheme of cooperative vision-based localization with communication constraints is proposed. Vehicles collect images of the environment and distance measurements between each other. Then vehicles transmit the coordinates of feature points and distances with constrained bits to the edge to estimate their positions. The \ac{fim} for absolute localization is first obtained, based on which we derive the relative \ac{speb} through subspace projection. Furthermore, we formulate the corresponding bit allocation problem for relative localization. Finally, a \ac{vgd} algorithm is developed by considering the influence of photographing, distance measurements and quantization noises. Compared with conventional bit allocation methods, numerical results demonstrate the localization performance gain of our proposed algorithm with higher computational efficiency.

eess.SY↗

arXiv2019

Effect of gallium doping on bubbling and helium retention in aluminum exposed to low-energy helium plasma

Surface bubbling and helium retention of pure aluminum and a solid solution of aluminum-gallium(Ga) alloy exposed to low-energy He plasma has been investigated with the fluence of 1.8E24 He/m2 at room temperature . Surface morphology observations show that after irradiation, the average size of He bubbles on Al-1.8 at.% Ga alloy is around 1.8 micro metre, smaller than that on pure Al. Ga doping noticeably increases the area density of the bubble. The thermal desorption spectroscopyshows that the release amount of He in Al-1.8 at.% Ga alloy is 1.77E21 He/m2, nearly three orders of magnitude higher than that in pure Al, whlie the He desorption peak of Al-1.8 at.% Ga alloy is 480 K, much lower than 580 K of pure Al. The results of slow positron annihilation spectroscopy (SPAS) indicate that the vacancy type defects were introduced by He plasma irradiation; and lattice distortion caused by Ga doping plays an important role in determining surface bubbling and He retention characteristics of Al-1.8 at.% Ga.

cond-mat.mtrl-sci↗