Two-photon polymerization-based 4D printing and its applications

Two-photon polymerization(TPP)is a cutting-edge micro/nanoscale three-dimensional(3D)printing technology based on the principle of two-photon absorption.TPP surpasses the diffraction limit in achieving feature sizes and excels in fabricating intricate 3D micro/nanostructures with exceptional resolution.The concept of 4D entails the fabrication of structures utilizing smart materials capable of undergoing shape,property,or functional changes in response to external stimuli over time.The integration of TPP and 4D printing introduces the possibility of producing responsive structures with micro/nanoscale accuracy,thereby enhancing the capabilities and potential applications of both technologies.This paper comprehensively reviews TPP-based 4D printing technology and its diverse applications.First,the working principles of TPP and its recent advancements are introduced.Second,the optional4D printing materials suitable for fabrication with TPP are discussed.Finally,this review paper highlights several noteworthy applications of TPP-based 4D printing,including domains such as biomedical microrobots,bioinspired microactuators,autonomous mobile microrobots,transformable devices and robots,as well as anti-counterfeiting microdevices.In conclusion,this paper provides valuable insights into the current status and future prospects of TPP-based4D printing technology,thereby serving as a guide for researchers and practitioners.

Projection micro stereolithography based 3D printing and its applications

Projection micro stereolithography(PμSL)is a high-resolution(up to 0.6μm)3D printing technology based on area projection triggered photopolymerization,and capable of fabricating complex 3D architectures covering multiple scales and with multiple materials.This paper reviews the recent development of the PμSL based 3D printing technologies,together with the related applications.It introduces the working principle,the commercialized products,and the recent multiscale,multimaterial printing capability of PμSL as well as some functional photopolymers that are suitable to PμSL.This review paper also summarizes a few typical applications of PμSL including mechanical metamaterials,optical components,4D printing,bioinspired materials and biomedical applications,and offers perspectives on the directions of the further development of PμSL based 3D printing technology.

How graph convolutions amplify popularity bias for recommendation?

Graphconvolutional networks(GCNs)have become prevalent in recommender system(RS)due to their superiority in modeling collaborative patterns.Although improving the overall accuracy,GCNs unfortunately amplify popularity bias-tail items are less likely to be recommended.This effect prevents the GCN-based RS from making precise and fair recommendations,decreasing the effectiveness of recommender systems in the long run.In this paper,we investigate how graph convolutions amplify the popularity bias in RS.Through theoretical analyses,we identify two fundamental factors:(1)with graph convolution(i.e.,neighborhood aggregation),popular items exert larger influence than tail items on neighbor users,making the users move towards popular items in the representation space;(2)after multiple times of graph convolution,popular items would affect more high-order neighbors and become more influential.The two points make popular items get closer to almost users and thus being recommended more frequently.To rectify this,we propose to estimate the amplified effect of popular nodes on each node's representation,and intervene the effect after each graph convolution.Specifically,we adopt clustering to discover highly-influential nodes and estimate the amplification effect of each node,then remove the effect from the node embeddings at each graph convolution layer.Our method is simple and generic-it can be used in the inference stage to correct existing models rather than training a new model from scratch,and can be applied to various GCN models.We demonstrate our method on two representative GCN backbones LightGCN and UltraGCN,verifying its ability in improving the recommendations of tail items without sacrificing the performance of popular items.Codes are open-sourced^(1)).

Stiffness-Tunable Origami Structures via Multimaterial Three-Dimensional Printing

Origami structure has been employed in many engineering applications.However,there is currently no strategy that can systematically achieve stiffness-tunable origami(STO)structures through proper geometric design.Here,we report a strategy for designing and fabricating STO structures based on thick-panel origami using multimaterial 3D printing.By adjusting the soft hinge position,we tune the geometric parameterψto program the stiffness and strength of origami structures.We develop origami structures with graded stiffness and strength by stacking Kresling origami structures with differentψ.The printed structures show great cyclic characteristics and deformation ability.After optimizing combinations of structures with differentψ,the multi-layer Kresling STO structures can effectively reduce the peak impact,showing a good energy absorption effect.The proposed approach can be implemented in various origami patterns to design and tune the mechanical properties of origami structures for many potential applications.

Rumor detection with self-supervised learning on texts and social graph

Rumor detection has become an emerging and active research field in recent years.At the core is to model the rumor characteristics inherent in rich information,such as propagation patterns in social network and semantic patterns in post content,and differentiate them from the truth.However,existing works on rumor detection fall short in modeling heterogeneous information,either using one single information source only(e.g.,social network,or post content)or ignoring the relations among multiple sources(e.g.,fusing social and content features via simple concatenation).Therefore,they possibly have drawbacks in comprehensively understanding the rumors,and detecting them accurately.In this work,we explore contrastive self-supervised learning on heterogeneous information sources,so as to reveal their relations and characterize rumors better.Technically,we supplement the main supervised task of detection with an auxiliary self-supervised task,which enriches post representations via post self-discrimination.Specifically,given two heterogeneous views of a post(i.e.,representations encoding social patterns and semantic patterns),the discrimination is done by maximizing the mutual information between different views of the same post compared to that of other posts.We devise cluster-wise and instance-wise approaches to generate the views and conduct the discrimination,considering different relations of information sources.We term this framework as self-supervised rumor detection(SRD).Extensive experiments on three real-world datasets validate the effectiveness of SRD for automatic rumor detection on social media.

Laser-based micro/nanofabrication in one, two and three dimensions

Advanced micro/nanofabrication of functional materials and structures with various dimensions represents a key research topic in modem nanoscience and technology and becomes critically important for numerous emerging technologies such as nanoelectronics, nanopho- tonics and micro/nanoelectromechanical systems. This review systematically explores the non-conventional material processing approaches in fabricating nanomaterials and micro/nanostructures of various dimensions which are challenging to be fabricated via conventional approaches. Research efforts are focused on laser-based techniques for the growth and fabrication of one-dimensional （1D）, two-dimensional （2D） and three-dimensional （3D） nanomaterials and micro/nanostructures. The following research topics are covered, including： 1） laser-assisted chemical vapor deposition （CVD） for highly efficient growth and integration of 1D nanomaterial of carbon nanotubes （CNTs）, 2） laser direct writing （LDW） of graphene ribbons under ambient conditions, and 3） LDW of 3D micro/nanostructures via additive and subtractive processes. Comparing with the conventional fabrication methods, the laser-based methods exhibit several unique advantages in the micro/nanofabrication of advanced functional materials and structures. For the 1D CNT growth, the laser-assisted CVD process can realize both rapid material synthesis and tight control of growth location and orientation of CNTs due to the highly intense energy delivery and laser-induced optical near-field effects. For the 2D graphene synthesis and patterning, roomtemperature and open-air fabrication of large-scale graphene patterns on dielectric surface has been successfully realized by a LDW process. For the 3D micro/nanofabrica- tion, the combination of additive two-photon polymeriza- tion （TPP） and subtractive multi-photon ablation （MPA） processes enables the fabrication of arbitrary complex 3D micro/nanostructures which tional fabrication methods are challenging for conven- Considering the numerous unique advantages of laser-based techniques, the laser- based micro/nanofabrication is expected to play a more and more important role in the fabrication of advanced functional micro/nano-devices.

Hydrogel-elastomer-based stretchable strain sensor fabricated by a simple projection lithography method

Stretchable strain sensor detects a wide range of strain variation and is therefore a key component in various applications.Unlike traditional ones made of elastomers doped with conductive components or fabricated with liquid conductors,ionically conductive hydrogel-based strain sensors remain conductive under large deformations and are biocompatible.However,dehydration is a challenging issue for the latter.Researchers have developed hydrogel-elastomer-based strain sensors where an elastomer matrix encapsulates a hydrogel circuit to prevent its dehydration.However,the reported multistep approaches are generally time-consuming.Our group recently reported a multimaterial 3D printing approach that enables fast fabrication of such sensors,yet requires a self-built digital-light-processing-based multimaterial 3D printer.Here,we report a simple projection lithography method to fabricate hydrogel-elastomer-based stretchable strain sensors within 5 minutes.This method only requires a UV projector/lamp with photomasks;the chemicals are commercially available;the protocols for preparing the polymer precursors are friendly to users without chemistry background.Moreover,the manufacturing flexibility allows users to readily pattern the sensor circuit and attach the sensor to a 3D printed soft pneumatic actuator to enable strain sensing on the latter.The proposed approach paves a simple and versatile way to fabricate hydrogel-elastomer-based stretchable strain sensors and flexible electronic devices.

Graph convolution machine for context-aware recommender system

The latest advance in recommendation shows that better user and item representations can be learned via performing graph convolutions on the user-item interaction graph.However,such finding is mostly restricted to the collaborative filtering(CF)scenario,where the interaction contexts are not available.In this work,we extend the advantages of graph convolutions to context-aware recommender system(CARS,which represents a generic type of models that can handle various side information).We propose Graph Convolution Machine(GCM),an end-to-end framework that consists of three components:an encoder,graph convolution(GC)layers,and a decoder.The encoder projects users,items,and contexts into embedding vectors,which are passed to the GC layers that refine user and item embeddings with context-aware graph convolutions on the user-item graph.The decoder digests the refined embeddings to output the prediction score by considering the interactions among user,item,and context embeddings.We conduct experiments on three real-world datasets from Yelp and Amazon,validating the effectiveness of GCM and the benefits of performing graph convolutions for CARS.

Dynamical behaviors of recurrently connected neural networks and linearly coupled networks with discontinuous right-hand sides

The aim of this paper is to provide a sys- tematic review on the framework to analyze dynamics in recurrently connected neural networks with discontinu- ous right-hand sides with a focus on the authors＇ works in the past three years. The concept of the Filippov so- lution is employed to define the solution of the neural network systems by transforming them to differential in- clusions. The theory of viability provides a tool to study the existence and uniqueness of the solution and the Lya- punov function （functional） approach is used to investi- gate the global stability and synchronization. More pre- cisely, we prove that the diagonal-dominant conditions guarantee the existence, uniqueness, and stability of a general class of integro-differential equations with （al- most） periodic self-inhibitions, interconnection weights, inputs, and delays. This model is rather general and in- cludes the well-known Hopfield neural networks, Cohen- Grossberg neural networks, and cellular neural networks as special cases. We extend the absolute stability anal- ysis of gradient-like neural network model by relaxing the analytic constraints so that they can be employed to solve optimization problem with non-smooth cost func- tions. Furthermore, we study the global synchronization problem of a class of linearly coupled neural network with discontinuous right-hand sides.