arXiv subjects

Feifan Wang

Publications and source records attributed to Feifan Wang.

At least 19 recordsLinked to original sources

arXiv2026

Curvature-Guided Sheaf Diffusion for Unsupervised Community Detection on Heterophilic Graphs

Detecting communities in heterophilic graphs -- where connected nodes often belong to different classes -- is hard for unsupervised methods: classical modularity and spectral methods are feature agnostic, while deep graph-clustering methods rely on contrastive or generative machinery that is opaque. We propose Curvature-Guided Sheaf Diffusion (CGSD), a fully unsupervised community-detection algorithm that uses the discrete Forman--Ricci curvature of each edge as its single topological signal, propagated through every stage of an end-to-end pipeline. CGSD makes three concrete contributions: (i)~a curvature-gated sheaf-diffusion encoder that gates edge messages by $\sigma(\kappa_e)$ and is trained from three label-free structural losses (modularity, anti-collapse, curvature-weighted reconstruction); (ii)~a curvature-aware spectral clusterer (CSpec) that re-weights the $k$-NN affinity of the embedding by $\sigma(\alpha \kappa_{e^*})$ before Ng--Jordan--Weiss; and (iii)~a unified label-free evaluation against nine truly-unsupervised baselines. On five heterophilic benchmarks (Cora, Cornell, Texas, Wisconsin, Chameleon), CGSD wins outright on Wisconsin and Chameleon and is competitive on the remaining three against nine unsupervised baselines. The gain over the strongest baseline is driven by the clusterer, not the encoder: on the same embedding, CSpec improves mean NMI from $0.091$ with $K$-Means to $0.107$ ($+15\%$, paired $t$-test $p=0.008$). The mechanism is interpretable: intra-community and inter-community curvature distributions are visibly separated. Code is open-sourced at https://github.com/woodywff/cgsd.

cs.LG↗

arXiv2026

Clownfish: Scaling DAG-based BFT Consensus via Sparse Edges

Directed Acyclic Graph (DAG) based BFT protocols have demonstrated the capability to achieve significantly high throughput in practice. Recent advancements focused on minimizing the good-case latency of these protocols, approaching the theoretical lower bound. However, the high communication complexity inherent in existing DAG-based protocols limits their scalability. This primarily arises because each vertex in the DAG must include a linear number of edges (references) to vertices from previous rounds. We present Clownfish, a partially synchronous DAG-based BFT protocol designed to address the scalability bottleneck. Clownfish achieves lower communication complexity by selectively reducing the number of edges in DAG vertices. When using a communication-optimal consistent broadcast, Clownfish attains quadratic total communication complexity per round, outperforming prior DAG-based protocols. Clownfish also reduces the additional latency in failure cases by optimizing the round advancement rule. Additionally, Clownfish supports multiple leaders per round to reduce average latency while maintaining its lower communication complexity. Our experimental evaluation demonstrates that Clownfish provides significantly better scalability than existing DAG-based protocols.

cs.DC↗

arXiv2025

Ultrafast {\mu}eV-Precision Bandgap Engineering in Low-Dimensional Topological Insulators

Precise and ultrafast control of electronic band structures is a central challenge for advancing quantum functional materials and devices. Conventional approaches--such as chemical doping, lattice strain, or external gating--offer robust stability but remain confined to the quasi-static regime, far from the intrinsic femto- to picosecond dynamics that govern many-body interactions. Here, using cryogenic transient reflectance spectroscopy, we realize dynamic bandgap engineering in the anisotropic topological insulator $\alpha$-Bi$_4$Br$_4$ with unprecedented micro-electron-volt ($\mu$eV) precision. The exceptional sensitivity arises from the cooperative action of long-lived topological carriers, stabilized by restricted bulk-to-edge scattering phase space, together with symmetry-resolved coherent phonons that modulate inter-chain hopping. These channels jointly modify Coulomb screening and interband transitions, enabling both gradual and oscillatory control of the electronic structure. Supported by first-principles and tight-binding theory, we further demonstrate a dual-pump coherent control strategy for continuous, mode-selective tuning of electronic energies with $\mu$eV accuracy. This framework paves the way for ultrafast on-demand band-structure engineering, pointing toward new frontiers in quantum optoelectronics, precision measurement in molecular and biological systems, and attosecond control of matter.

cond-mat.mes-hall↗

arXiv2025

Optical multistability in a compact microcavity enabled by near-exceptional coupling

Multistability -- the emergence of multiple stable states under identical conditions -- is a hallmark of nonlinear complexity and an enabling mechanism for multilevel optical memory and photonic computing. Its realization in a compact footprint, however, is limited by intrinsically weak optical nonlinearities and the enlarged free spectral range that raises the multistability threshold. Here, we overcome this constraint by engineering a pair of spectrally close, ultra-high-Q resonances in a photonic crystal microcavity. Leveraging structural perturbations that deliberately introduce non-Hermitian coupling through a shared radiation channel, we drive the resonances toward an exceptional point with nearly degenerate wavelengths and balanced quality factors approaching $10^6$. This configuration substantially enhances thermo-optical nonlinearity and produces pronounced tristability and hysteresis loops within a footprint of 20 {\mu}m at input powers below 240 {\mu}W. We further demonstrate proof-of-concept optical random-access memory through controlled switching among multistable states. These results establish a general strategy for nonlinear microcavities to achieve energy-efficient multistability for reconfigurable all-optical memories, logic, and neuromorphic processors.

physics.optics↗

arXiv2025

EmbodiedBrain: Expanding Performance Boundaries of Task Planning for Embodied Intelligence

The realization of Artificial General Intelligence (AGI) necessitates Embodied AI agents capable of robust spatial perception, effective task planning, and adaptive execution in physical environments. However, current large language models (LLMs) and multimodal LLMs (MLLMs) for embodied tasks suffer from key limitations, including a significant gap between model design and agent requirements, an unavoidable trade-off between real-time latency and performance, and the use of unauthentic, offline evaluation metrics. To address these challenges, we propose EmbodiedBrain, a novel vision-language foundation model available in both 7B and 32B parameter sizes. Our framework features an agent-aligned data structure and employs a powerful training methodology that integrates large-scale Supervised Fine-Tuning (SFT) with Step-Augumented Group Relative Policy Optimization (Step-GRPO), which boosts long-horizon task success by integrating preceding steps as Guided Precursors. Furthermore, we incorporate a comprehensive reward system, including a Generative Reward Model (GRM) accelerated at the infrastructure level, to improve training efficiency. For enable thorough validation, we establish a three-part evaluation system encompassing General, Planning, and End-to-End Simulation Benchmarks, highlighted by the proposal and open-sourcing of a novel, challenging simulation environment. Experimental results demonstrate that EmbodiedBrain achieves superior performance across all metrics, establishing a new state-of-the-art for embodied foundation models. Towards paving the way for the next generation of generalist embodied agents, we open-source all of our data, model weight, and evaluating methods, which are available at https://zterobot.github.io/EmbodiedBrain.github.io.

cs.CV↗

arXiv2025

Wave morphing and flat-top ground states in photonics systems driven by artificial gauge fields

In quantum physics, classical optics, and many other wave systems, wave confinement in a potential well is associated with discrete oscillatory states, and the ground state is typically assumed to vanish uniformly. An open question is whether the ground state can counterintuitively support a flat-top, nonzero envelope, offering new opportunities for quantum emitters, optical antennas, and lasers. Here, we show that by applying the Byers-Yang theorem with an artificial gauge field, energy levels can be continuously shifted, driving eigenstates to morph into a ground state with a uniform yet nontrivial wave envelope. We implement this concept in a photonic crystal slab where a central bulk region is surrounded by heterogeneous bandgaps that engineer reflective phases acting as an artificial local gauge field. By inducing lasing, we probe directly the evolution of the energy levels, demonstrating wave morphing toward a flat-top ground state via near- and far-field measurements.

physics.optics↗

arXiv2025

Emotion Knowledge Enhancement for Vision Large Language Models: A Self-Verification Approach for High-Quality Emotion Instruction Data Generation

Facial emotion perception in the vision large language model (VLLM) is crucial for achieving natural human-machine interaction. However, creating high-quality annotations for both coarse- and fine-grained facial emotion analysis demands costly expertise. The lack of such high-quality instruction data limits the performance of VLLMs in facial emotion perception. To address this, we propose a self-verification approach with emotion knowledge enhancement (SEKE), which generates high-quality instruction data for multi-grained emotion analysis cost-effectively using closed-source VLLM. This approach integrates prior human knowledge to VLLM inference, guided by the inherent correlations between three grained levels of emotion descriptions, i.e., discrete expression, valence-arousal, and action unit, to reliably generate comprehensive annotations. A self-verification strategy with Uncertainty-Aware Monte Carlo sampling (SV-UAMC) is further embedded to efficiently extract more accurate VLLM predictions, further improving annotation reliability. Consequently, we construct a facial emotion instruction dataset (FEID) containing three comprehensive descriptions, which provides coarse- and fine-grained emotional information for effective model training. Additionally, we introduce a facial emotion analysis benchmark (FEAB) to measure the VLLM's corresponding ability. Our method significantly outperforms state-of-the-art methods on three downstream facial emotion analysis tasks.

cs.LG↗

arXiv2025

Optimizing LLM Inference Throughput via Memory-aware and SLA-constrained Dynamic Batching

The increasing adoption of large language models (LLMs) necessitates inference serving systems that can deliver both high throughput and low latency. Deploying LLMs with hundreds of billions of parameters on memory-constrained GPUs exposes significant limitations in static batching methods. Current inference serving systems often treat batch sizes as fixed hyper-parameters, hindering real-time adaptation to varying system conditions. In this paper, we propose a dynamic batching method that continuously monitors memory utilization and adheres to service-level agreements (SLAs) to enable real-time batch size configuration adjustment. The method comprises two core components: a memory-aware batch scheduler that dynamically allocates GPU resources and a latency feedback mechanism that optimizes decoding processes under SLA constraints. The numerical experiments demonstrate throughput gains of 8% to 28% and capacity improvements of 22% compared to traditional static batching methods, while maintaining full compatibility with existing inference infrastructure. These results highlight the effectiveness of dynamic batching in balancing computational efficiency and quality-of-service requirements for contemporary LLM deployment scenarios. The source code of this work is publicly available at https://github.com/KevinLee1110/dynamic-batching.

cs.DC↗

arXiv2025

Hybrid Offline-online Scheduling Method for Large Language Model Inference Optimization

With the development of large language models (LLMs), it has become increasingly important to optimize hardware usage and improve throughput. In this paper, we study the inference optimization of the serving system that deploys LLMs. To optimize system throughput and maximize hardware utilization, we formulate the inference optimization problem as a mixed-integer programming (MIP) model and propose a hybrid offline-online method as solution. The offline method improves large-scale inference systems by introducing a Minimizing Makespan Bin Packing Problem. We further provide a theoretical lower bound computation method. Then, we propose an online sorting and preemptive scheduling method to better utilize hardware. In the online iteration scheduling process, a Lagrangian method is applied to evaluate the cost efficiency of inserting prefill stages versus decode stages at each iteration and dynamically determine when to preempt decoding tasks and insert prefill tasks. Experiments using real-world data from the LLaMA-65B model and the GSM8K dataset demonstrate that system utilization improves from 80.2% to 89.1%, and the total inference time decreases from 201.00 to 190.58 seconds. A 100-cases study shows that our method consistently outperforms the baseline method and improves the utilization rate by 8.0% on average. Finally, we discuss potential future extensions, including stochastic modeling, reinforcement learning-based schedulers, and dynamic decision-making strategies for system throughput and hardware utilization.

cs.DC↗

arXiv2025

Structural phase transitions between layered Indium Selenide for inte-grated photonic memory

The primary mechanism of optical memristive devices relies on the phase transitions between amorphous-crystalline states. The slow or energy hungry amorphous-crystalline transitions in optical phase-change materials are detrimental to the devices scalability and performance. Leveraging the integrated photonic platform, we demonstrate a single nanosecond pulse triggered nonvolatile and reversible switching between two layered structures of indium selenide (In2Se3). High resolution pair distribution function reveals the detailed atomistic transition pathways between the layered structures. With inter-layer shear glide and isosymmetric phase transition, the switching between alpha and beta structural states contain low re-configurational entropy, allowing reversible switching between layered structures. Broadband refractive index contrast, optical transparency, and volumetric effect in the crystalline-crystalline phase transition are experimentally characterized in molecular beam epitaxy-grown thin films and compared to ab initials calculations. The nonlinear resonator transmission spectra measure an incremental linear loss rate of 3.3 GHz introduced by 1.5 micrometer long In2Se3 covered lay-er, resulting from the combinations of material absorption and scattering.

physics.optics↗

arXiv2024

Miniaturized spectrometer enabled by end-to-end deep learning on large-scale radiative cavity array

Miniaturized (mini-) spectrometers are highly desirable tools for chemical, biological, and medical diagnostics because of their potential for portable and in situ spectral detection. In this work, we propose and demonstrate a mini-spectrometer that combines a large-scale radiative cavity array with end-to-end deep learning networks. Specifically, we utilize high-Q bound states in continuum cavities with distinct radiation characteristics as the fundamental units to achieve parallel spectral detection. We realize a 36 $\times$ 30 cavity array that spans a wide spectral range from 1525 to 1605 nm with quality factors above 10^4. We further train a deep network with 8000 outputs to directly map arbitrary spectra to array responses excited by the out-of-plane incident. Experimental results demonstrate that the proposed mini-spectrometer can resolve unknown spectra with a resolution of 0.048 nm in a bandwidth of 80 nm and fidelity exceeding 95%, thus offering a promising method for compact, high resolution, and broadband spectroscopy.

physics.optics↗

arXiv2024

From a Bird's Eye View to See: Joint Camera and Subject Registration without the Camera Calibration

We tackle a new problem of multi-view camera and subject registration in the bird's eye view (BEV) without pre-given camera calibration. This is a very challenging problem since its only input is several RGB images from different first-person views (FPVs) for a multi-person scene, without the BEV image and the calibration of the FPVs, while the output is a unified plane with the localization and orientation of both the subjects and cameras in a BEV. We propose an end-to-end framework solving this problem, whose main idea can be divided into following parts: i) creating a view-transform subject detection module to transform the FPV to a virtual BEV including localization and orientation of each pedestrian, ii) deriving a geometric transformation based method to estimate camera localization and view direction, i.e., the camera registration in a unified BEV, iii) making use of spatial and appearance information to aggregate the subjects into the unified BEV. We collect a new large-scale synthetic dataset with rich annotations for evaluation. The experimental results show the remarkable effectiveness of our proposed method.

cs.CV↗

arXiv2024

Unveiling the Power of Self-supervision for Multi-view Multi-human Association and Tracking

Multi-view multi-human association and tracking (MvMHAT), is a new but important problem for multi-person scene video surveillance, aiming to track a group of people over time in each view, as well as to identify the same person across different views at the same time, which is different from previous MOT and multi-camera MOT tasks only considering the over-time human tracking. This way, the videos for MvMHAT require more complex annotations while containing more information for self learning. In this work, we tackle this problem with a self-supervised learning aware end-to-end network. Specifically, we propose to take advantage of the spatial-temporal self-consistency rationale by considering three properties of reflexivity, symmetry and transitivity. Besides the reflexivity property that naturally holds, we design the self-supervised learning losses based on the properties of symmetry and transitivity, for both appearance feature learning and assignment matrix optimization, to associate the multiple humans over time and across views. Furthermore, to promote the research on MvMHAT, we build two new large-scale benchmarks for the network training and testing of different algorithms. Extensive experiments on the proposed benchmarks verify the effectiveness of our method. We have released the benchmark and code to the public.

cs.CV↗

arXiv2023

Ultra-low-loss optical interconnect enabled by topological unidirectional guided resonance

Grating couplers that interconnect photonic chips to off-chip components are of essential importance for various optoelectronics applications. Despite numerous efforts in past decades, existing grating couplers still suffer from poor energy efficiency and thus hinder photonic integration toward a larger scale. Here, we theoretically propose and experimentally demonstrate a method to achieve ultra-low-loss grating coupler by employing topological unidirectional guided resonances (UGRs). Leveraging the unidirectional emitting nature of UGRs, the useless downward radiation is greatly suppressed with no mirror placed on the bottom. By engineering the dispersion and apodizing the geometry of grating, we realize a grating coupler on 340 nm silicon-on-insulator platform with a record-low-loss of 0.34 dB and bandwidth exceeding 30 nm at the telecom wavelength of 1550 nm. We further show a pair of grating couplers works as optic via that interconnects two stacked photonic chips with a loss of only 0.94 dB. Our work sheds light on the feasibility of energy-efficient optical interconnect for silicon photonics, and paving the way to large-scale photonic integration for applications from optical communication to photonic computing.

physics.optics↗

arXiv2023

Photocatalytic oxidation in few-layer Tellurene for loss-invariant integrated photonic resonance trimming

Two-dimensional materials with unique physicochemical properties promote photocatalytic activities. As the 2D material composites research studies the statistical average of complex catalytic behaviors, an integrated photonic platform allows clean and single flake level photo-catalytic investigations with precisely quantified photocatalytic activities. In this paper, we track fluence-dependent photo-oxidation in two-dimensional Tellurene (2D Te) by the evanescently coupled micro-resonator. Nearly 32 perent of oxidation is achieved in 10 nm 2D Te flake, compared to only 4.5% oxidation in 30 nm sample, probed by the resonance shift in silicon microring resonators (MRRs) substrate. The wider bandgap in the few layers of 2D Te allows faster charge transfer to adsorbed oxygen for a more efficient photocatalytic redox reaction. The photo-oxidation in hybrid 2D Te results in an invariant lineshapes of optical transmission resonance for wavelength trimming (more than 3 times resonance bandwidth). The low threshold power, near-infrared, and in-waveguide resonance trimming scheme is compatible with most integrated photonic setups for easy fixing the nanofabrication-induced random resonance deviation for integrated photonic circuit applications in wavelength-division-multiplexing systems and spin qubits quantum computing.

physics.optics↗

arXiv2023

Scalable photonic crystal waveguides with 2 dB component loss

Periodic or gradient subwavelength structures are basic configurations of photonic crystals and metamaterials. The measured linear losses of those nanophotonic devices are well-beyond theoretical predictions. Nanofabrication related geometric inhomogeneity is considered as the primary cause of the deleterious performance. The deep-UV photolithography in CMOS foundry is a large-scale parallel processing, which can significantly suppress the random offsets and thus the optical linear loss. Here we demonstrate ultra-low loss photonic crystal waveguides with a multi-project wafer run through AIM photonics and post-processing. For sub-millimeter long photonic crystal waveguides, 2 dB total loss and 40 dB extinction ratio are observed across dies.

physics.optics↗

arXiv2023

Chiral exceptional point and coherent suppression of backscattering in silicon microring with low loss Mie scatterer

Non-Hermitian systems with their spectral degeneracies known as exceptional points (EPs) have been explored for lasing, controlling light transport, and enhancing a sensor s response. A ring resonator can be brought to an EP by controlling the coupling between its frequency degenerate clockwise and counterclockwise traveling modes. This has been typically achieved by introducing two or more nanotips into the resonator s mode volume. While this method provides a route to study EP physics, the basic understanding of how the nanotips shape and size symmetry impact the system s non-Hermicity is missing, along with additional loss from both in-plane and out-of-plane scattering. The limited resonance stability poses a challenge for leveraging EP effects for switches or modulators, which requires stable cavity resonance and fixed laser-cavity detuning. Here we use lithographically defined asymmetric and symmetric Mie scatterers, which enable subwavelength control of wave transmission and reflections without deflecting to additional radiation channels. We show that those pre-defined Mie scatterers can bring the system to an EP without post tuning, as well as enable chiral light transport within the resonator. Counterintuitively, the Mie scatterer results in enhanced quality factor measured on the transmission port, through coherently suppressing the backscattering from the waveguide surface roughness. The proposed device platform enables pre-defined chiral light propagation and backscattering-free resonances, needed for various applications such as frequency combs, solitons, sensing, and other nonlinear optical processes such as photon blockade, and regenerative oscillators.

physics.optics↗

arXiv2023

Slow light silicon modulator beyond 110 GHz bandwidth

Silicon modulators are key components in silicon photonics to support the dense integration of electro-optic (EO) functional elements on a compact chip for various applications including high-speed data transmission, signal processing, and photonic computing. Despite numerous advances in promoting the operation speed of silicon modulators, a bandwidth ceiling of 67 GHz emerges in practices and becomes an obstacle to paving silicon photonics toward Tbps level data throughput on a single chip. Here, we theoretically propose and experimentally demonstrate a design strategy for silicon modulators by employing the slow light effect, which shatters the present bandwidth ceiling of silicon modulators and pushes its limit beyond 110 GHz in a small footprint. The proposed silicon modulator is built on a coupled-resonator optical waveguide (CROW) architecture, in which a set of Bragg gratings are appropriately cascaded to give rise to a slow light effect. By comprehensively balancing a series of merits including the group index, photon lifetime, electrical bandwidth, and losses, we found the modulators can benefit from the slow light for better modulation efficiency and compact size while remaining their bandwidth sufficiently high to support ultra-high-speed data transmission. Consequently, we realize a modulator with an EO bandwidth of 110 GHz in a length of 124 μm, and demonstrate a data rate beyond 110 Gbps by applying simple on-off keying modulation for a DSP-free operation. Our work proves that silicon modulators beyond 110 GHz are feasible, thus shedding light on the potentials of silicon photonics in ultra-high-bandwidth applications such as data communication, optical interconnection, and photonic machine learning.

physics.optics↗