arXiv subjects

Jingcheng Li

Publications and source records attributed to Jingcheng Li.

At least 19 recordsLinked to original sources

arXiv2026

Tunable Flat Bands and magnetism in Triangulene-based Superatomic Graphene

Superatomic graphene platforms host a rich portfolio of flat-band-driven exotic quantum properties, yet their experimental realization remains challenging. Here, we report the bottom-up on-surface synthesis of superatomic graphene using phosphorus-doped triangulene as building blocks. Scanning tunneling microscopy and spectroscopy measurements resolve the well-defined honeycomb lattice of as-fabricated superatomic graphene and demonstrate the characteristic Dirac band and flat band electronic structures. Density functional theory calculations reveal that the flat bands originate from the in-plane p$_x,_y$-like frontier orbitals of the phosphorus-doped triangulene units, leading to intrinsic half-metallic behavior. Furthermore, oxygen functionalization of the molecular precursor enables deterministic modulation of the electronic structure and magnetic ordering. This work establishes a general platform for designing correlated quantum materials with tunable flat band properties.

cond-mat.mtrl-sci↗

arXiv2025

Exotic Surface Stripe Orders in Correlated Kagome Metal CsCr3Sb5

The newly discovered kagome superconductor CsCr3Sb5 exhibits distinct features with flat bands and unique magnetism, providing a compelling platform for exploring novel quantum states of correlated electron systems. Emergent charge order in this material is a key for understanding unconventional superconductivity, but it remains unexplored at the atomic scale and the underlying physics is elusive. Here, we identify and unreported stripe orders on the surface which are distinct from the bulk and investigate the underlying bulk electronic properties using a combination of scanning tunneling microscopy (STM), angle-resolved photoemission spectroscopy (ARPES) and density functional theory (DFT) calculations. Specifically, a mixture of 2a0 * a0 and 3a0 * a0 stripe order is found on Cs-terminated surface while 4a0 * root3a0 stripe order is found on the Sb-terminated surface. The electronic spectra exhibit strongly correlated features resembling that of high temperature superconductors, with kagome flat bands lying about 330 meV above EF, suggesting that the electron correlations arise from Coulomb interactions and Hund's coupling. Moreover, a distinct electron-boson coupling mode is observed at approximately 100 meV. These findings provide new insights into the interplay between surface and bulk charge orders in this strongly correlated kagome system.

cond-mat.str-el↗

arXiv2025

Unifying Language Agent Algorithms with Graph-based Orchestration Engine for Reproducible Agent Research

Language agents powered by large language models (LLMs) have demonstrated remarkable capabilities in understanding, reasoning, and executing complex tasks. However, developing robust agents presents significant challenges: substantial engineering overhead, lack of standardized components, and insufficient evaluation frameworks for fair comparison. We introduce Agent Graph-based Orchestration for Reasoning and Assessment (AGORA), a flexible and extensible framework that addresses these challenges through three key contributions: (1) a modular architecture with a graph-based workflow engine, efficient memory management, and clean component abstraction; (2) a comprehensive suite of reusable agent algorithms implementing state-of-the-art reasoning approaches; and (3) a rigorous evaluation framework enabling systematic comparison across multiple dimensions. Through extensive experiments on mathematical reasoning and multimodal tasks, we evaluate various agent algorithms across different LLMs, revealing important insights about their relative strengths and applicability. Our results demonstrate that while sophisticated reasoning approaches can enhance agent capabilities, simpler methods like Chain-of-Thought often exhibit robust performance with significantly lower computational overhead. AGORA not only simplifies language agent development but also establishes a foundation for reproducible agent research through standardized evaluation protocols.

cs.CL↗

arXiv2025

Epitaxial growth of gold films on the elemental superconductors V(100), Nb(100) and Nb(110)

Quantum technologies require a new generation of superconducting electronic devices and circuitry. However, the superconducting materials used to construct them are restricted to a class of bulk superconductors. Gold films grown in contact with superconducting materials can exhibit superconducting correlations through the proximity effect, with various possible implementations in quantum technology. Here, we study the growth of flat Au films on various surfaces of the elemental superconductors vanadium and niobium through a combination of low-temperature scanning tunneling microscopy (STM) and X-ray photoelectron spectroscopy (XPS). In particular, we investigate the growth morphology and composition as a function of temperature and coverage. We find that gold films can grow flat and oxygen-free when annealed to sufficiently high temperatures; however, they are also susceptible to partial intermixing with the substrate elements. Low-temperature scanning tunneling spectroscopy (STS) measurements elucidate the emergence of a proximitized superconducting gap at the gold surface. Additionally, we demonstrate the survival of the magnetic state of FeTPP-Cl molecules on the surface of these gold films, proving that they behave as a good support for probing molecular magnetism. We found that the exchange interaction between the molecular spin and the superconducting condensate can be inferred by the measurement of sub-gap Yu-Shiva-Rusinov states. Our work shows that proximitised Au films constitute a promising platform to explore on-superconducting-surface synthesis and the interaction between superconductivity and magnetism in a large spin-orbit coupling environment.

cond-mat.supr-con↗

arXiv2025

VLM-R1: A Stable and Generalizable R1-style Large Vision-Language Model

Recently DeepSeek R1 has shown that reinforcement learning (RL) can substantially improve the reasoning capabilities of Large Language Models (LLMs) through a simple yet effective design. The core of R1 lies in its rule-based reward formulation, which leverages tasks with deterministic ground-truth answers to enable precise and stable reward computation. In the visual domain, we similarly observe that a wide range of visual understanding tasks are inherently equipped with well-defined ground-truth annotations. This property makes them naturally compatible with rule-based reward mechanisms. Motivated by this observation, we investigate the extension of R1-style reinforcement learning to Vision-Language Models (VLMs), aiming to enhance their visual reasoning capabilities. To this end, we develop VLM-R1, a dedicated framework designed to harness RL for improving VLMs' performance on general vision-language tasks. Using this framework, we further explore the feasibility of applying RL to visual domain. Experimental results indicate that the RL-based model not only delivers competitive performance on visual understanding tasks but also surpasses Supervised Fine-Tuning (SFT) in generalization ability. Furthermore, we conduct comprehensive ablation studies that uncover a series of noteworthy insights, including the presence of reward hacking in object detection, the emergence of the "OD aha moment", the impact of training data quality, and the scaling behavior of RL across different model sizes. Through these analyses, we aim to deepen the understanding of how reinforcement learning enhances the capabilities of vision-language models, and we hope our findings and open-source contributions will support continued progress in the vision-language RL community. Our code and model are available at https://github.com/om-ai-lab/VLM-R1

cs.CV↗

arXiv2024

Half-Metallicity in Triangulene-based Superatomic Graphene

The discovery of two-dimensional (2D) magnets has opened up new possibilities for miniaturizing spintronic devices to the monolayer limit. 2D half-metals, capable of conducting fully spin-polarized currents when spin-orbit coupling is minimal, provide a key advantage in improving device performance. Extensive theoretical research has been carried out to discover 2D half-metals, yet their realization remains elusive. Here we report the bottom-up synthesis of superatomic graphene and the demonstration of its half-metallic properties. The produced graphene half-metal is fabricated through an on-surface synthetic approach with phosphorus-doped triangulene as its building block. Scanning tunneling microscopy measurements reveal its metallic band structures and identify its ferromagnetism through magnon excitation under varying magnetic fields. Density functional theory simulations accurately capture its half-metallic characteristics, uncovering the origin of spin-polarized bands from the p$_x,_y$-like orbital of superatomic graphene. Our work demonstrates intrinsic 2D carbon magnetism, paving the way for harnessing its advantages in spintronics.

cond-mat.mes-hall↗

arXiv2024

RSEND: Retinex-based Squeeze and Excitation Network with Dark Region Detection for Efficient Low Light Image Enhancement

Images captured under low-light scenarios often suffer from low quality. Previous CNN-based deep learning methods often involve using Retinex theory. Nevertheless, most of them cannot perform well in more complicated datasets like LOL-v2 while consuming too much computational resources. Besides, some of these methods require sophisticated training at different stages, making the procedure even more time-consuming and tedious. In this paper, we propose a more accurate, concise, and one-stage Retinex theory based framework, RSEND. RSEND first divides the low-light image into the illumination map and reflectance map, then captures the important details in the illumination map and performs light enhancement. After this step, it refines the enhanced gray-scale image and does element-wise matrix multiplication with the reflectance map. By denoising the output it has from the previous step, it obtains the final result. In all the steps, RSEND utilizes Squeeze and Excitation network to better capture the details. Comprehensive quantitative and qualitative experiments show that our Efficient Retinex model significantly outperforms other CNN-based models, achieving a PSNR improvement ranging from 0.44 dB to 4.2 dB in different datasets and even outperforms transformer-based models in the LOL-v2-real dataset.

cs.CV↗

arXiv2024

TG-NAS: Generalizable Zero-Cost Proxies with Operator Description Embedding and Graph Learning for Efficient Neural Architecture Search

Neural Architecture Search (NAS) is a powerful technique for discovering high-performing CNN architectures, but most existing methods rely on costly training or extensive sampling. Zero-shot NAS offers a training-free alternative by using proxies to predict architecture performance. However, existing proxies are often suboptimal -- frequently outperformed by simple metrics like parameter count or FLOPs -- and they generalize poorly across different search spaces. Moreover, current model-based proxies struggle to adapt to new operators without access to ground-truth accuracy, limiting their transferability. We propose TG-NAS, a universal, model-based zero-cost (ZC) proxy that combines a Transformer-based operator embedding generator with a Graph Convolutional Network (GCN) to predict architecture performance. Unlike prior model-based predictors, TG-NAS requires no retraining and generalizes across arbitrary search spaces. It serves as a standalone ZC proxy with strong data efficiency, robustness, and cross-space consistency. Extensive evaluations across diverse NAS benchmarks demonstrate TG-NAS's superior rank correlation and generalizability compared to existing proxies. Additionally, it improves search efficiency by up to 300x and discovers architectures achieving 93.75% CIFAR-10 accuracy on NAS-Bench-201 and 74.9% ImageNet top-1 accuracy on the DARTS space, establishing TG-NAS as a promising foundation for efficient, generalizable NAS.

cs.LG↗

arXiv2024

Tunable Current Rectification Through a Designer Graphene Nanoribbon

Unimolecular current rectifiers are fundamental building blocks in organic electronics. Rectifying behavior has been identified in numerous organic systems due to electron-hole asymmetries of orbital levels interfaced by a metal electrode. As a consequence, the rectifying ratio (RR) determining the diode efficiency remains fixed for a chosen molecule-metal interface. Here, we present a mechanically tunable molecular diode exhibiting an exceptionally large rectification ratio (>10^5) and reversible direction. The molecular system comprises a 7-armchair graphene nanoribbon (GNR) doped with a single unit of substitutional diboron within its structure, synthesized with atomic precision on a gold substrate by on-surface synthesis. The diboron unit creates half-populated in-gap bound states and splits the GNR frontier bands into two segments, localizing the bound state in a double barrier configuration. By suspending these GNRs freely between the tip of a low-temperature scanning tunneling microscope and the substrate, we demonstrate unipolar hole transport through the boron in-gap state's resonance. Strong current rectification is observed, associated with the varying widths of the two barriers, which can be tuned by altering the distance between tip and substrate. This study introduces an innovative approach for the precise manipulation of molecular electronic functionalities, opening new avenues for advanced applications in organic electronics.

cond-mat.mes-hall↗

arXiv2023

Unbiased organism-agnostic and highly sensitive signal peptide predictor with deep protein language model

Signal peptide (SP) is a short peptide located in the N-terminus of proteins. It is essential to target and transfer transmembrane and secreted proteins to correct positions. Compared with traditional experimental methods to identify signal peptides, computational methods are faster and more efficient, which are more practical for analyzing thousands or even millions of protein sequences, especially for metagenomic data. Here we present Unbiased Organism-agnostic Signal Peptide Network (USPNet), a signal peptide classification and cleavage site prediction deep learning method that takes advantage of protein language models. We propose to apply label distribution-aware margin loss to handle data imbalance problems and use evolutionary information of protein to enrich representation and overcome species information dependence.

cs.AI↗

arXiv2023

SecBeam: Securing mmWave Beam Alignment against Beam-Stealing Attacks

Millimeter wave (mmWave) communications employ narrow-beam directional communications to compensate for the high path loss at mmWave frequencies. Compared to their omnidirectional counterparts, an additional step of aligning the transmitter's and receiver's antennas is required. In current standards such as 802.11ad, this beam alignment process is implemented via an exhaustive search through the horizontal plane known as beam sweeping. However, the beam sweeping process is unauthenticated. As a result, an adversary, Mallory, can launch an active beam-stealing attack by injecting forged beacons of high power, forcing the legitimate devices to beamform towards her direction. Mallory is now in control of the communication link between the two devices, thus breaking the false sense of security given by the directionality of mmWave transmissions. Prior works have added integrity protection to beam alignment messages to prevent forgeries. In this paper, we demonstrate a new beam-stealing attack that does not require message forging. We show that Mallory can amplify and relay a beam sweeping frame from her direction without altering its contents. Intuitively, cryptographic primitives cannot verify physical properties such as the SNR used in beam selection. We propose a new beam sweeping protocol called SecBeam that utilizes power/sector randomization and coarse angle-of-arrival information to detect amplify-and-relay attacks. We demonstrate the security and performance of SecBeam using an experimental mmWave platform and via ray-tracing simulations.

cs.CR↗

arXiv2023

Secret-Free Device Pairing in the mmWave Band

Many Next Generation (NextG) applications feature devices that are capable of communicating and sensing in the Millimeter-Wave (mmWave) bands. Trust establishment is an important first step to bootstrap secure mmWave communication links, which is challenging due to the lack of prior secrets and the fact that traditional cryptographic authentication methods cannot bind digital trust with physical properties. Previously, context-based device pairing approaches were proposed to extract shared secrets from common context, using various sensing modalities. However, they suffer from various limitations in practicality and security. In this work, we propose the first secret-free device pairing scheme in the mmWave band that explores the unique physical-layer properties of mmWave communications. Our basic idea is to let Alice and Bob derive common randomness by sampling physical activity in the surrounding environment that disturbs their wireless channel. They construct reliable fingerprints of the activity by extracting event timing information from the channel state. We further propose an uncoordinated path hopping mechanism to resolve the challenges of beam alignment for activity sensing without prior trust. A key novelty of our protocol is that it remains secure against both co-located passive adversaries and active Man-in-the-Middle attacks, which is not possible with existing context-based pairing approaches. We implement our protocol in a 28GHz mmWave testbed, and experimentally evaluate its security in realistic indoor environments. Results show that our protocol can effectively thwart several different types of adversaries.

cs.CR↗

arXiv2023

PoF: Proof-of-Following for Vehicle Platoons

Cooperative vehicle platooning significantly improves highway safety, fuel efficiency, and traffic flow. In this model, a set of vehicles move in line formation and coordinate acceleration, braking, and steering using a combination of physical sensing and vehicle-to-vehicle (V2V) messaging. The authenticity and integrity of the V2V messages are paramount to safety. For this reason, recent V2V and V2X standards support the integration of a PKI. However, a PKI cannot bind a vehicle's digital identity to the vehicle's physical state (location, velocity, etc.). As a result, a vehicle with valid cryptographic credentials can impact platoons from a remote location. In this paper, we seek to provide the missing link between the physical and the digital world in the context of vehicle platooning. We propose a new access control protocol we call Proof-of-Following (PoF) that verifies the following distance between a candidate and a verifier. The main idea is to draw security from the common, but constantly changing environment experienced by the closely traveling vehicles. We use the large-scale fading effect of ambient RF signals as a common source of randomness to construct a {\em PoF} primitive. The correlation of large-scale fading is an ideal candidate for the mobile outdoor environment because it exponentially decays with distance and time. We evaluate our PoF protocol on an experimental platoon of two vehicles in freeway, highway, and urban driving conditions. We demonstrate that the PoF withstands both the pre-recording and following attacks with overwhelming probability.

cs.CR↗

arXiv2023

Distilled Mid-Fusion Transformer Networks for Multi-Modal Human Activity Recognition

Human Activity Recognition is an important task in many human-computer collaborative scenarios, whilst having various practical applications. Although uni-modal approaches have been extensively studied, they suffer from data quality and require modality-specific feature engineering, thus not being robust and effective enough for real-world deployment. By utilizing various sensors, Multi-modal Human Activity Recognition could utilize the complementary information to build models that can generalize well. While deep learning methods have shown promising results, their potential in extracting salient multi-modal spatial-temporal features and better fusing complementary information has not been fully explored. Also, reducing the complexity of the multi-modal approach for edge deployment is another problem yet to resolve. To resolve the issues, a knowledge distillation-based Multi-modal Mid-Fusion approach, DMFT, is proposed to conduct informative feature extraction and fusion to resolve the Multi-modal Human Activity Recognition task efficiently. DMFT first encodes the multi-modal input data into a unified representation. Then the DMFT teacher model applies an attentive multi-modal spatial-temporal transformer module that extracts the salient spatial-temporal features. A temporal mid-fusion module is also proposed to further fuse the temporal features. Then the knowledge distillation method is applied to transfer the learned representation from the teacher model to a simpler DMFT student model, which consists of a lite version of the multi-modal spatial-temporal transformer module, to produce the results. Evaluation of DMFT was conducted on two public multi-modal human activity recognition datasets with various state-of-the-art approaches. The experimental results demonstrate that the model achieves competitive performance in terms of effectiveness, scalability, and robustness.

cs.HC↗

arXiv2022

Wiggle: Physical Challenge-Response Verification of Vehicle Platooning

Autonomous vehicle platooning promises many benefits such as fuel efficiency, road safety, reduced traffic congestion, and passenger comfort. Platooning vehicles travel in a single file, in close distance, and at the same velocity. The platoon formation is autonomously maintained by a Cooperative Adaptive Cruise Control (CACC) system which relies on sensory data and vehicle-to-vehicle (V2V) communications. In fact, V2V messages play a critical role in shortening the platooning distance while maintaining safety. Whereas V2V message integrity and source authentication can be verified via cryptographic methods, establishing the truthfulness of the message contents is a much harder task. This work establishes a physical access control mechanism to restrict V2V messages to platooning members. Specifically, we aim at tying the digital identity of a candidate requesting to join a platoon to its physical trajectory relative to the platoon. We propose the {\em Wiggle} protocol that employs a physical challenge-response exchange to prove that a candidate requesting to be admitted into a platoon actually follows it. The protocol name is inspired by the random longitudinal movements that the candidate is challenged to execute. {\em Wiggle} prevents any remote adversary from joining the platoon and injecting fake CACC messages. Compared to prior works, {\em Wiggle} is resistant to pre-recording attacks and can verify that the candidate is directly behind the verifier at the same lane.

cs.CR↗

arXiv2022

Addressing electron spins embedded in metallic graphene nanoribbons

Spin-hosting graphene nanostructures are promising metal-free systems for elementary quantum spintronic devices. Conventionally, spins are protected from quenching by electronic bandgaps, which also hinder electronic access to their quantum state. Here, we present a narrow graphene nanoribbon substitutionally doped with boron heteroatoms that combines a metallic character with the presence of localized spin 1/2 states in its interior. The ribbon was fabricated by on-surface synthesis on a Au(111) substrate. Transport measurements through ribbons suspended between the tip and the sample of a scanning tunnelling microscope revealed their ballistic behavior, characteristic of metallic nanowires. Conductance spectra show fingerprints of localized spin states in form of Kondo resonances and inelastic tunnelling excitations. Density functional theory rationalizes the metallic character of the graphene nanoribbon due to the partial depopulation of the valence band induced by the boron atoms. The transferred charge builds localized magnetic moments around the boron atoms. The orthogonal symmetry of the spin-hosting state's and the valence band's wavefunctions protects them from mixing, maintaining the spin states localized. The combination of ballistic transport and spin localization into a single graphene nanoribbon offers the perspective of electronically addressing and controlling carbon spins in real device architectures.

cond-mat.mes-hall↗

arXiv2022

Extending the spin excitation lifetime of a magnetic molecule on a proximitized superconductor

Magnetic molecules deposited on surfaces are a promising platform to individually address and manipulate spins. Long spin excitation lifetimes are necessary to utilize them in quantum information processing and data storage. Normally, coupling of the molecular spin with the conduction electrons of metallic surfaces causes fast relaxation of spin excitations into the ground state. However, the presence of superconducting paring effects in the substrate can protect the excited spin from decaying. In this work, we show that a proximity-induced superconducting gold film can sustain spin excitations of a FeTPP-Cl molecule for more than 80ns. This long value was determined by studying inelastic spin excitations of the S=5/2 multiplet of FeTPP-Cl on Au films over V(100) using scanning tunneling spectroscopy. The spin lifetime decreases with increasing film thickness, in apparent connection with the gradual gap-closing of a pair of de Gennes-Saint James resonances found inside the superconducting gap. Our results elucidate the use of proximitized gold electrodes for addressing quantum spins on surfaces, envisioning new routes for tuning the value of their spin lifetime.

cond-mat.mes-hall↗

arXiv2021

Topological phase transition in chiral graphene nanoribbons: from edge bands to end states

Precise control over the size and shape of graphene nanostructures allows engineering spin-polarized edge and topological states, representing a novel source of non-conventional $π$-magnetism with promising applications in quantum spintronics. A prerequisite for their emergence is the existence of robust gapped phases, which are difficult to find in extended graphene systems: only armchair graphene nanoribbons (GNRs) show a band gap that, however, closes for any other GNR orientation. Here we show that semi-metallic chiral GNRs (chGNRs) narrowed down to nanometer widths undergoes a topological phase transition, becoming first topological insulators, and transforming then into trivial band insulators for the narrowest chGNRs. We fabricated atomically precise chGNRs of different chirality and size by on surface synthesis using predesigned molecular precursors. Combining scanning tunnelling microscopy (STM) measurements and theory simulations, we follow the evolution of topological properties and bulk band gap depending on the width, length, and chirality of chGNRs. The first emerging gapped phases are topological, protected by a chiral interaction pattern between edges. For narrower ribbons, the symmetry of the interaction pattern changes, and the topological gap closes and re-opens again as a trivial band insulator. Our findings represent a new platform for producing topologically protected spin states and demonstrates the potential of connecting chiral edge and defect structure with band engineering.

cond-mat.mes-hall↗