arXiv subjects

Jialin Li

Publications and source records attributed to Jialin Li.

At least 73 records · Page 4Linked to original sources

arXiv2025

Postselected amplification and photon recycling applied to optical sensing of magnetic fields

We apply the combined technique of postselected amplification and photon-recycling to an optical setup of magnetic field precision measurement. We propose two recycling schemes and carry out analytic expressions for the amplified signal and measurement sensitivity. The results show significant improvement of performance over conventional measurement. The underlying reason is twofold. On one aspect, introducing the technique of recycling eliminates the shortcoming of data discarding in postselection, thus maintains similar noise level of conventional measurement (without postselection). On the other aspect, performing intentional postselection within the recycling framework, which was originally proposed in the context of gravitational wave detection, can amplify the signal. Thus, the measurement signal-to-noise ratio is enhanced.

quant-ph↗

arXiv2025

Jacobi forms of weight one on $Γ_0(N)$

Let $J_{1,m}(N)$ be the vector space of Jacobi forms of weight one and index $m$ on $Γ_0(N)$. In 1985, Skoruppa proved that $J_{1,m}(1)=0$ for all $m$. In 2007, Ibukiyama and Skoruppa proved that $J_{1,m}(N)=0$ for all $m$ and all squarefree $N$ with $\mathrm{gcd}(m,N)=1$. This paper aims to extend their results. We determine all levels $N$ separately, such that $J_{1,m}(N)=0$ for all $m$; or $J_{1,m}(N)=0$ for all $m$ with $\mathrm{gcd}(m,N)=1$. We also establish explicit dimension formulas of $J_{1,m}(N)$ when $m$ and $N$ are relatively prime or $m$ is squarefree. These results are obtained by refining Skoruppa's method and analyzing local invariants of Weil representations. As applications, we prove the vanishing of Siegel modular forms of degree two and weight one in some cases.

math.NT↗

arXiv2025

In-Loop Filtering Using Learned Look-Up Tables for Video Coding

In-loop filtering (ILF) is a key technology in video coding standards to reduce artifacts and enhance visual quality. Recently, neural network-based ILF schemes have achieved remarkable coding gains, emerging as a powerful candidate for next-generation video coding standards. However, the use of deep neural networks (DNN) brings significant computational and time complexity or high demands for dedicated hardware, making it challenging for general use. To address this limitation, we study a practical ILF solution by adopting look-up tables (LUTs). After training a DNN with a restricted reference range for ILF, all possible inputs are traversed, and the output values of the DNN are cached into LUTs. During the coding process, the filtering process is performed by simply retrieving the filtered pixel through locating the input pixels and interpolating between the cached values, instead of relying on heavy inference computations. In this paper, we propose a universal LUT-based ILF framework, termed LUT-ILF++. First, we introduce the cooperation of multiple kinds of filtering LUTs and propose a series of customized indexing mechanisms to enable better filtering reference perception with limited storage consumption. Second, we propose the cross-component indexing mechanism to enable the filtering of different color components jointly. Third, in order to make our solution practical for coding uses, we propose the LUT compaction scheme to enable the LUT pruning, achieving a lower storage cost of the entire solution. The proposed framework is implemented in the VVC reference software. Experimental results show that the proposed framework achieves on average 0.82%/2.97%/1.63% and 0.85%/4.11%/2.06% bitrate reduction for common test sequences, under the AI and RA configurations, respectively. Compared to DNN-based solutions, our proposed solution has much lower time complexity and storage cost.

eess.IV↗

arXiv2025

Two-sided-loop jet originates from the filament internal reconnection

Magnetic reconnection driving two-sided-loop jet is typically associated with interactions between an emerging bipole and the overlying horizontal magnetic field, or between filaments from separate magnetic systems. Leveraging high temporal and spatial resolution observations from ground-based and space-borne instruments, we have identified a two-sided-loop jet originating from magnetic reconnection between threads within a single filament. Our observations show that as two initially crossing filamentary threads within the filament converge, reconnection takes place at their intersection. In the Doppler images, distinct redshift and blueshift signals are observed at the locations where the filament threads intersected. This process generates a two-sided-loop jet with outflow speeds of \speed{22.2} and \speed{62.5}. Following reconnection, the original crossing threads transform into two parallel threads that subsequently separate at speeds of \speed{2.8} and \speed{8.3}. This observation offers a new perspective on the mechanisms responsible for jet formation.

astro-ph.SR↗

arXiv2025

Wide Separation Planets In Time (WISPIT): Discovery of a Gap H$α$ Protoplanet WISPIT 2b with MagAO-X

Excellent (<25 mas) H$_α$ images of the star TYC 5709-354-1 led to the discovery of a rare H$_α$ protoplanet. This star was discovered by the WISPIT survey to have a large multi-ring transitional disk, and is hereafter WISPIT 2. Our H$_α$ images of 2025, April 13 and April 16 discovered an accreting (H$_α$ in emission) protoplanet: WISPIT 2b (r=309.43$\pm$1.56 mas; (~54 au deprojected), PA=242.21$\pm$0.41 degrees) likely clearing a dust-free gap between the two brightest dust rings in the transitional disk. Our SNR=12.5 detection gave an H$_α$ ASDI contrast of (6.5$\pm$0.5)x10$^{-4}$ and a H$_α$ line flux of (1.29$\pm$0.28)x10$^{-15}$ erg/s/cm$^2$. We also present L' photometry from LBT/LMIRcam of the planet (L'=15.30$\pm$0.05 mag) which, when coupled with an age of 5.1$^{+2.4}_{-1.3}$ Myr, yields a planet mass estimate of 5.3$\pm$1.0 Mjup from the DUSTY evolutionary models. WISPIT 2b is accreting at 2.25$^{-0.17}_{+3.75}$x10$^{-12}$ Msun/yr. WISPIT 2b is very similar to the other H$_α$ protoplanets in terms of mass, age, flux, and accretion rate. The inclination of the system (${\it i}$=44 degrees) is also, surprisingly, very similar to the other known H$α$ protoplanet systems which all cluster from 37$\leq{\it i}\leq$52 degrees. We argue this clustering has only a ~1.0% (2.6 sigma) probability of occurring randomly, and so we speculate that magnetospherical accretion might have a preferred inclination range (~37-52 degrees) for the direct (cloud free, low extinction) line of sight to the H-alpha line formation/shock region. We also find at 110mas (~15au deprojected) a close companion candidate (CC1) which may be consistent with an inner dusty 9$\pm$4 Mjup planet.

astro-ph.EP↗

arXiv2025

Discovery of H$α$ Emission from a Protoplanet Candidate Around the Young Star 2MASS J16120668-3010270 with MagAO-X

2MASS J16120668-3010270 (hereafter 2MJ1612) is a young M0 star that hosts a protoplanetary disk in the Upper Scorpious star-forming region. Recent ALMA observations of 2MJ1612 show a mildly inclined disk ($i$=37$^\circ$) with a large dust-depleted gap (R$_\text{cav}\approx$0.4" or 53 au). We present high-contrast H$α$ observations from MagAO-X on the 6.5m Magellan Telescope and new high resolution sub-mm dust continuum observations with ALMA of 2MJ1612. On both 2025 April 13 and 16, we recovered a point source with H$α$ excess with SNR $\gtrsim$5 within the disk gap in our MagAO-X Angular and Spectral Differential (ASDI) images at a separation of 141.96$\pm$2.10 mas (23.45$\pm$0.29 au deprojected) from the star and position angle (PA)= 159.00$\pm$0.55$^\circ$. Furthermore, this H$α$ source is within close proximity to a K band point source in SPHERE/IRDIS observation taken on 2023 July 21 \citep{sphere2025sub}. The astrometric offset between the K band and H$α$ source can be explained by orbital motion of a bound companion. Thus our observations can be best explained by the discovery of an accreting protoplanet, 2MJ1612 b, with an estimated mass of 4$M_\text{Jup}$ and H$α$ line flux ranging from (29.7 $\pm$7.5)$\times$10$^{-16}$ ergs/s/cm$^2$ to (8.2$\pm$3.4)$\times$10$^{-16}$ ergs/s/cm$^2$. 2MJ1612 b is likely the third example of an accreting H$α$ protoplanet responsible for carving the gap in its host disk, joining PDS 70b and c. Further study is necessary to confirm and characterize this protoplanet candidate and to identify any additional protoplanets that may also play a role in shaping the gap.

astro-ph.EP↗

arXiv2025

A CLIP-based Uncertainty Modal Modeling (UMM) Framework for Pedestrian Re-Identification in Autonomous Driving

Re-Identification (ReID) is a critical technology in intelligent perception systems, especially within autonomous driving, where onboard cameras must identify pedestrians across views and time in real-time to support safe navigation and trajectory prediction. However, the presence of uncertain or missing input modalities--such as RGB, infrared, sketches, or textual descriptions--poses significant challenges to conventional ReID approaches. While large-scale pre-trained models offer strong multimodal semantic modeling capabilities, their computational overhead limits practical deployment in resource-constrained environments. To address these challenges, we propose a lightweight Uncertainty Modal Modeling (UMM) framework, which integrates a multimodal token mapper, synthetic modality augmentation strategy, and cross-modal cue interactive learner. Together, these components enable unified feature representation, mitigate the impact of missing modalities, and extract complementary information across different data types. Additionally, UMM leverages CLIP's vision-language alignment ability to fuse multimodal inputs efficiently without extensive finetuning. Experimental results demonstrate that UMM achieves strong robustness, generalization, and computational efficiency under uncertain modality conditions, offering a scalable and practical solution for pedestrian re-identification in autonomous driving scenarios.

cs.CV↗

arXiv2025

Refining Critical Thinking in LLM Code Generation: A Faulty Premise-based Evaluation Framework

With the advancement of code generation capabilities in large language models (LLMs), their reliance on input premises has intensified. When users provide inputs containing faulty premises, the probability of code generation hallucinations rises significantly, exposing deficiencies in their self-scrutiny capabilities. This paper proposes Faulty Premises Bench (FPBench), the first code generation evaluation framework targeting faulty premises. By systematically constructing three categories of faulty premises and integrating multi-dimensional evaluation metrics, it conducts in-depth assessments of 15 representative LLMs. The key findings are as follows: (1) Most models exhibit poor reasoning abilities and suboptimal code generation performance under faulty premises, heavily relying on explicit prompts for error detection, with limited self-scrutiny capabilities; (2) Faulty premises trigger a point of diminishing returns in resource investment, leading to blindly increasing length fails to enhance quality; (3) The three types of faulty premises respectively activate distinct defect patterns in models, revealing a triple dissociation in the cognitive mechanisms of code generation models. This study not only highlights the urgent need for LLMs to proactively verify premises in code generation but also, through the proposed FPBench framework and multi-dimensional evaluation system, provides a theoretical foundation and practical pathway for developing reliable, human-centric code generation models.

cs.AI↗

arXiv2025

TAPAS: Fast and Automatic Derivation of Tensor Parallel Strategies for Large Neural Networks

Tensor parallelism is an essential technique for distributed training of large neural networks. However, automatically determining an optimal tensor parallel strategy is challenging due to the gigantic search space, which grows exponentially with model size and tensor dimension. This prohibits the adoption of auto-parallel systems on larger models. We observe that neural networks usually contain repeated substructures, and build an automatic parallelism framework named TAPAS that eliminates redundant search efforts. TAPAS employs a divide-and-conquer approach that efficiently folds the search space by identifying those unique substructures. As a result, it runs at sub-linear complexity concerning the model size, making it a scalable solution for training large-scale networks. Our evaluations demonstrate that TAPAS outperforms the state-of-the-art automatic parallelism frameworks by up to $160\times$ in search speed on a wide range of models, and the performance of derived strategies is competitive or even better compared with the expert-engineered Megatron-LM library.

cs.LG↗

arXiv2025

RoboGSim: A Real2Sim2Real Robotic Gaussian Splatting Simulator

Efficient acquisition of real-world embodied data has been increasingly critical. However, large-scale demonstrations captured by remote operation tend to take extremely high costs and fail to scale up the data size in an efficient manner. Sampling the episodes under a simulated environment is a promising way for large-scale collection while existing simulators fail to high-fidelity modeling on texture and physics. To address these limitations, we introduce the RoboGSim, a real2sim2real robotic simulator, powered by 3D Gaussian Splatting and the physics engine. RoboGSim mainly includes four parts: Gaussian Reconstructor, Digital Twins Builder, Scene Composer, and Interactive Engine. It can synthesize the simulated data with novel views, objects, trajectories, and scenes. RoboGSim also provides an online, reproducible, and safe evaluation for different manipulation policies. The real2sim and sim2real transfer experiments show a high consistency in the texture and physics. We compared the test results of RoboGSim data and real robot data on both RoboGSim and real robot platforms. The experimental results show that the RoboGSim data model can achieve zero-shot performance on the real robot, with results comparable to real robot data. Additionally, in experiments with novel perspectives and novel scenes, the RoboGSim data model performed even better on the real robot than the real robot data model. This not only helps reduce the sim2real gap but also addresses the limitations of real robot data collection, such as its single-source and high cost. We hope RoboGSim serves as a closed-loop simulator for fair comparison on policy learning. More information can be found on our project page https://robogsim.github.io/.

cs.RO↗

arXiv2025

Building State Machine Replication Using Practical Network Synchrony

Distributed systems, such as state machine replication, are critical infrastructures for modern applications. Practical distributed protocols make minimum assumptions about the underlying network: They typically assume a partially synchronous or fully asynchronous network model. In this work, we argue that modern data center systems can be designed to provide strong synchrony properties in the common case, where servers move in synchronous lock-step rounds. We prove this hypothesis by engineering a practical design that uses a combination of kernel-bypass network, multithreaded architecture, and loosened round length, achieving a tight round bound under 2us. Leveraging our engineered networks with strong synchrony, we co-design a new replication protocol, Chora. Chora exploits the network synchrony property to efficiently pipeline multiple replication instances, while allowing all replicas to propose in parallel without extra coordination. Through experiments, we show that Chora achieves 255% and 109% improvement in throughput over state-of-the-art single-leader and multi-leader protocols, respectively.

cs.DC↗

arXiv2025

PipeOffload: Improving Scalability of Pipeline Parallelism with Memory Optimization

Pipeline parallelism (PP) is widely used for training large language models (LLMs), yet its scalability is often constrained by high activation memory consumption as the number of in-flight microbatches grows with the degree of PP. In this paper, we focus on addressing this challenge by leveraging the under-explored memory offload strategy in PP. With empirical study, we discover that in the majority of standard configurations, at least half, and potentially all, of the activations can be offloaded with negligible overhead. In the cases where full overload is not possible, we introduce a novel selective offload strategy that decreases peak activation memory in a better-than-linear manner. Furthermore, we integrate memory offload with other techniques to jointly consider overall throughput and memory limitation. Our experiments proves that the per-device activation memory effectively reduces with the total number of stages, making PP a stronger alternative than TP, offering up to a 19\% acceleration with even lower memory consumption. The implementation is open-sourced at \href{https://github.com/sail-sg/zero-bubble-pipeline-parallelism}{this url}.

cs.LG↗

arXiv2025

On the vanishing order of Jacobi forms at infinity

In this paper, we establish two types of upper bounds on the vanishing order of Jacobi forms at infinity. The first type is for classical Jacobi forms, which is optimal in a certain sense. The second type is for Jacobi forms of lattice index. Based on this bound, we obtain a lower bound on the slope of orthogonal modular forms, and we prove that the module of symmetric formal Fourier--Jacobi series on $\mathrm{O}(m,2)$ has finite rank.

math.NT↗

arXiv2025

Sums of four polygonal numbers: precise formulas

In this paper we give unified formulas for the numbers of representations of positive integers as sums of four generalized $m$-gonal numbers, and as restricted sums of four squares under a linear condition, respectively. These formulas are given as $\mathbb{Z}$-linear combinations of Hurwitz class numbers. As applications, we prove several Zhi-Wei Sun's conjectures. As by-products, we obtain formulas for expressing the Fourier coefficients of $\vartheta(τ,z)^4$, $η(τ)^{12}$, $η(τ)^4$ and $η(τ)^8η(2τ)^8$ in terms of Hurwitz class numbers, respectively. The proof is based on the theory of Jacobi forms.

math.NT↗

arXiv2025

AgentGroupChat-V2: Divide-and-Conquer Is What LLM-Based Multi-Agent System Need

Large language model based multi-agent systems have demonstrated significant potential in social simulation and complex task resolution domains. However, current frameworks face critical challenges in system architecture design, cross-domain generalizability, and performance guarantees, particularly as task complexity and number of agents increases. We introduces AgentGroupChat-V2, a novel framework addressing these challenges through three core innovations: (1) a divide-and-conquer fully parallel architecture that decomposes user queries into hierarchical task forest structures enabling dependency management and distributed concurrent processing. (2) an adaptive collaboration engine that dynamically selects heterogeneous LLM combinations and interaction modes based on task characteristics. (3) agent organization optimization strategies combining divide-and-conquer approaches for efficient problem decomposition. Extensive experiments demonstrate AgentGroupChat-V2's superior performance across diverse domains, achieving 91.50% accuracy on GSM8K (exceeding the best baseline by 5.6 percentage points), 30.4% accuracy on competition-level AIME (nearly doubling other methods), and 79.20% pass@1 on HumanEval. Performance advantages become increasingly pronounced with higher task difficulty, particularly on Level 5 MATH problems where improvements exceed 11 percentage points compared to state-of-the-art baselines. These results confirm that AgentGroupChat-V2 provides a comprehensive solution for building efficient, general-purpose LLM multi-agent systems with significant advantages in complex reasoning scenarios. Code is available at https://github.com/MikeGu721/AgentGroupChat-V2.

cs.CL↗

arXiv2025

DynaServe: Unified and Elastic Execution for Dynamic Disaggregated LLM Serving

LLM inference must meet strict latency SLOs (e.g., 100 ms P99 time-between-tokens) while maximizing goodput. Yet, real-world variability in prompt and response lengths skews compute-intensive prefill and memory-bound decode phases, making both colocated (even with chunked prefill) and disaggregated deployments unable to simultaneously deliver low tail latency and high throughput. We introduce DynaServe, a high-performance LLM serving system built atop vLLM that unifies and extends both paradigms for maximizing goodput under SLO constraints, when handling unbalanced and dynamic workloads. It relies on a micro-request abstraction, which arbitrarily splits each request at any token boundary into at most two cooperating segments. A two-level scheduling framework then balances micro-request load across unified GPU instances. The global scheduler rapidly selects per-request split points by considering both the request's prefill/decode time ratio and the current load across GPU instances. The local schedulers on each GPU instance independently form SLO-aware batches, adjusting their composition in response to workload fluctuations, potential latency spikes and per-GPU under/over utilization. On real-world traces, DynaServe boosts the overall serving capacity from 1.15$\times$ to 3.07$\times$, improves goodput by up to 1.91$\times$ and 1.61$\times$, and improves the performance by up to 60\% in a hybrid workload under SLO compared to state-of-the-art colocated and disaggregated baselines.

cs.DC↗

arXiv2025

Five New Sirius-Like White Dwarf + Main Sequence Star Systems with MagAO-X

Most known white dwarfs in multiple systems with main sequence stars have been discovered with M-type companions, because the white dwarf causes detectable UV excess and bluer colors than expected from a single M star. Surveys have shown that the number of white dwarfs in Sirius-like systems within 100 pc of the Sun is lower than expected, suggesting that white dwarfs are being missed in the glare of their main sequence companions. In this work we have leveraged the angular resolution and high-contrast capabilities, as well as optimization for visible wavelengths, of the extreme adaptive optics instrument MagAO-X to detect new white dwarf companions to AFGK stars. We present the first results of our survey with the extreme AO instrument MagAO-X, called the Pup Search, of 18 targets with seven new candidate companions, five of which are confirmed to be white dwarfs. We discuss the new detections in the context of previous surveys and other detection metric sensitivities and show that we are sensitive to a region not probed by other surveys. Finally we discuss the future of the Pup Search in light of developing technologies.

astro-ph.SR↗

arXiv2025

P3P: Pseudo-3D Pre-training for Scaling 3D Voxel-based Masked Autoencoders

3D pre-training is crucial to 3D perception tasks. Nevertheless, limited by the difficulties in collecting clean and complete 3D data, 3D pre-training has persistently faced data scaling challenges. In this work, we introduce a novel self-supervised pre-training framework that incorporates millions of images into 3D pre-training corpora by leveraging a large depth estimation model. New pre-training corpora encounter new challenges in representation ability and embedding efficiency of models. Previous pre-training methods rely on farthest point sampling and k-nearest neighbors to embed a fixed number of 3D tokens. However, these approaches prove inadequate when it comes to embedding millions of samples that feature a diverse range of point numbers, spanning from 1,000 to 100,000. In contrast, we propose a tokenizer with linear-time complexity, which enables the efficient embedding of a flexible number of tokens. Accordingly, a new 3D reconstruction target is proposed to cooperate with our 3D tokenizer. Our method achieves state-of-the-art performance in 3D classification, few-shot learning, and 3D segmentation. Code is available at https://github.com/XuechaoChen/P3P-MAE.

cs.CV↗