arXiv subjects

Jun Yang

Publications and source records attributed to Jun Yang.

At least 127 records · Page 7Linked to original sources

arXiv2025

Efficient Agent: Optimizing Planning Capability for Multimodal Retrieval Augmented Generation

Multimodal Retrieval-Augmented Generation (mRAG) has emerged as a promising solution to address the temporal limitations of Multimodal Large Language Models (MLLMs) in real-world scenarios like news analysis and trending topics. However, existing approaches often suffer from rigid retrieval strategies and under-utilization of visual information. To bridge this gap, we propose E-Agent, an agent framework featuring two key innovations: a mRAG planner trained to dynamically orchestrate multimodal tools based on contextual reasoning, and a task executor employing tool-aware execution sequencing to implement optimized mRAG workflows. E-Agent adopts a one-time mRAG planning strategy that enables efficient information retrieval while minimizing redundant tool invocations. To rigorously assess the planning capabilities of mRAG systems, we introduce the Real-World mRAG Planning (RemPlan) benchmark. This novel benchmark contains both retrieval-dependent and retrieval-independent question types, systematically annotated with essential retrieval tools required for each instance. The benchmark's explicit mRAG planning annotations and diverse question design enhance its practical relevance by simulating real-world scenarios requiring dynamic mRAG decisions. Experiments across RemPlan and three established benchmarks demonstrate E-Agent's superiority: 13% accuracy gain over state-of-the-art mRAG methods while reducing redundant searches by 37%.

cs.AI↗

arXiv2025

Spectral Hardening Reveals Afterglow Emergence in Long-Duration Fast X-ray Transients: A Case Study of GRB 250404A/EP250404a

The prompt emission and afterglow phases of gamma-ray bursts (GRBs) have been extensively studied, yet the transition between these two phases remains inadequately characterized due to limited multiwavelength observational coverage. Among the recent growing samples of fast X-ray transients observed by Einstein Probe (EP), a subgroup of GRBs are captured with long-duration X-ray emission, potentially containing featured evolution from prompt emission to the afterglow phase. In this Letter, we present a detailed analysis of GRB 250404A/EP250404a, a bright fast X-ray transient detected simultaneously by EP and the Fermi Gamma-ray Burst Monitor in X-rays and gamma rays. Its continuous X-ray emission reveals a long-duration tail, accompanied by distinct spectral evolution manifested by the spectral index $α_{\rm X}$ with an initial softening, followed by an evident hardening, eventually reaching a plateau at the value of $\sim$ -2. Early optical and near-infrared observations enable broadband modeling with forward- and reverse-shock components, confirming that the X-ray hardening signals the emergence of the external-shock afterglow. From this spectral hardening we infer that the prompt phase in soft X-rays lasted $\sim300\;\mathrm{s}$, which is more than 3 times longer than the gamma-ray $T_{90}$. This well-tracked soft-hard-flat spectral pattern provides a clear indication of afterglow emergence from the fading prompt emission and offers a practical criterion for identifying a distinct population of GRBs among fast X-ray transients, even when the detection of the gamma-ray counterpart or obvious temporal break is absent.

astro-ph.HE↗

arXiv2025

LLM-based Multi-Agent Copilot for Quantum Sensor

Large language models (LLM) exhibit broad utility but face limitations in quantum sensor development, stemming from interdisciplinary knowledge barriers and involving complex optimization processes. Here we present QCopilot, an LLM-based multi-agent framework integrating external knowledge access, active learning, and uncertainty quantification for quantum sensor design and diagnosis. Comprising commercial LLMs with few-shot prompt engineering and vector knowledge base, QCopilot employs specialized agents to adaptively select optimization methods, automate modeling analysis, and independently perform problem diagnosis. Applying QCopilot to atom cooling experiments, we generated 10${}^{\rm{8}}$ sub-$\rmμ$K atoms without any human intervention within a few hours, representing $\sim$100$\times$ speedup over manual experimentation. Notably, by continuously accumulating prior knowledge and enabling dynamic modeling, QCopilot can autonomously identify anomalous parameters in multi-parameter experimental settings. Our work reduces barriers to large-scale quantum sensor deployment and readily extends to other quantum information systems.

quant-ph↗

arXiv2025

Enhancing Robustness of Control Barrier Function: A Reciprocal Resistance-based Approach

In this note, a new reciprocal resistance-based control barrier function (RRCBF) is developed to enhance the robustness of control barrier functions for disturbed affine nonlinear systems, without requiring explicit knowledge of disturbance bounds. By integrating a reciprocal resistance-like term into the conventional zeroing barrier function framework, we formally establish the concept of the reciprocal resistance-based barrier function (RRBF), rigorously proving the forward invariance of its associated safe set and its robustness against bounded disturbances. The RRBF inherently generates a buffer zone near the boundary of the safe set, effectively dominating the influence of uncertainties and external disturbances. This foundational concept is extended to formulate RRCBFs, including their high-order variants. To alleviate conservatism in the presence of complex, time-varying disturbances, we further introduce a disturbance observer-based RRCBF (DO-RRCBF), which exploits disturbance estimates to enhance safety guarantees and recover nominal control performance. The effectiveness of the proposed framework is validated through two simulation studies: a second-order linear system illustrating forward invariance in the phase plane, and an adaptive cruise control scenario demonstrating robustness in systems with high relative degree.

eess.SY↗

arXiv2025

MMCircuitEval: A Comprehensive Multimodal Circuit-Focused Benchmark for Evaluating LLMs

The emergence of multimodal large language models (MLLMs) presents promising opportunities for automation and enhancement in Electronic Design Automation (EDA). However, comprehensively evaluating these models in circuit design remains challenging due to the narrow scope of existing benchmarks. To bridge this gap, we introduce MMCircuitEval, the first multimodal benchmark specifically designed to assess MLLM performance comprehensively across diverse EDA tasks. MMCircuitEval comprises 3614 meticulously curated question-answer (QA) pairs spanning digital and analog circuits across critical EDA stages - ranging from general knowledge and specifications to front-end and back-end design. Derived from textbooks, technical question banks, datasheets, and real-world documentation, each QA pair undergoes rigorous expert review for accuracy and relevance. Our benchmark uniquely categorizes questions by design stage, circuit type, tested abilities (knowledge, comprehension, reasoning, computation), and difficulty level, enabling detailed analysis of model capabilities and limitations. Extensive evaluations reveal significant performance gaps among existing LLMs, particularly in back-end design and complex computations, highlighting the critical need for targeted training datasets and modeling approaches. MMCircuitEval provides a foundational resource for advancing MLLMs in EDA, facilitating their integration into real-world circuit design workflows. Our benchmark is available at https://github.com/cure-lab/MMCircuitEval.

cs.LG↗

arXiv2025

DiffCkt: A Diffusion Model-Based Hybrid Neural Network Framework for Automatic Transistor-Level Generation of Analog Circuits

Analog circuit design consists of the pre-layout and layout phases. Among them, the pre-layout phase directly decides the final circuit performance, but heavily depends on experienced engineers to do manual design according to specific application scenarios. To overcome these challenges and automate the analog circuit pre-layout design phase, we introduce DiffCkt: a diffusion model-based hybrid neural network framework for the automatic transistor-level generation of analog circuits, which can directly generate corresponding circuit structures and device parameters tailored to specific performance requirements. To more accurately quantify the efficiency of circuits generated by DiffCkt, we introduce the Circuit Generation Efficiency Index (CGEI), which is determined by both the figure of merit (FOM) of a single generated circuit and the time consumed. Compared with relative research, DiffCkt has improved CGEI by a factor of $2.21 \sim 8365\times$, reaching a state-of-the-art (SOTA) level. In conclusion, this work shows that the diffusion model has the remarkable ability to learn and generate analog circuit structures and device parameters, providing a revolutionary method for automating the pre-layout design of analog circuits. The circuit dataset will be open source, its preview version is available at https://github.com/CjLiu-NJU/DiffCkt.

cs.ET↗

arXiv2025

AnalogTester: A Large Language Model-Based Framework for Automatic Testbench Generation in Analog Circuit Design

Recent advancements have demonstrated the significant potential of large language models (LLMs) in analog circuit design. Nevertheless, testbench construction for analog circuits remains manual, creating a critical bottleneck in achieving fully automated design processes. Particularly when replicating circuit designs from academic papers, manual Testbench construction demands time-intensive implementation and frequent adjustments, which fails to address the dynamic diversity and flexibility requirements for automation. AnalogTester tackles automated analog design challenges through an LLM-powered pipeline: a) domain-knowledge integration, b) paper information extraction, c) simulation scheme synthesis, and d) testbench code generation with Tsinghua Electronic Design (TED). AnalogTester has demonstrated automated Testbench generation capabilities for three fundamental analog circuit types: operational amplifiers (op-amps), bandgap references (BGRs), and low-dropout regulators (LDOs), while maintaining a scalable framework for adaptation to broader circuit topologies. Furthermore, AnalogTester can generate circuit knowledge data and TED code corpus, establishing fundamental training datasets for LLM specialization in analog circuit design automation.

cs.MA↗

arXiv2025

Ocean Tides on Asynchronously Rotating Planets Orbiting Low-mass Stars

Planets in the liquid-water habitable zone of low-mass stars experience large tidal forces, $10^3$ to $10^4$ times those on Earth, due to the small distance between the habitable zone and the host stars. Therefore, interior solid tides, ocean tides and atmospheric tides on these planets could be much stronger than that on Earth, but rare work has been done to explicitly simulate the ocean tides. Here, for the first time, we perform global ocean tide simulations and show that ocean tides on asynchronously rotating planets with large eccentricities can reach $\mathcal{O}(1000)\,\mathrm{m}$ in height and $\mathcal{O}(10)\,\mathrm{m\,s^{-1}}$ in flow speed. Interactions between tide and bottom topography can induce large energy dissipation, $\sim\mathcal{O}(100)\,\mathrm{W\,m^{-2}}$ in global mean. This tidal energy dissipation can strongly accelerate orbital evolution by 1-2 orders of magnitude. However, for planets with small eccentricities, the ocean tides are much weaker but still comparable to that on modern Earth. Our results suggest that ocean tides on eccentric planets orbiting low-mass stars are orders of magnitude more powerful than those on Earth and can dramatically influence surface geography and orbital evolution.

astro-ph.EP↗

arXiv2025

ReviewInstruct: A Review-Driven Multi-Turn Conversations Generation Method for Large Language Models

The effectiveness of large language models (LLMs) in conversational AI is hindered by their reliance on single-turn supervised fine-tuning (SFT) data, which limits contextual coherence in multi-turn dialogues. Existing methods for generating multi-turn dialogue data struggle to ensure both diversity and quality in instructions. To address this, we propose Review-Instruct, a novel framework that synthesizes multi-turn conversations through an iterative "Ask-Respond-Review" process involving three agent roles: a Candidate, multiple Reviewers, and a Chairman. The framework iteratively refines instructions by incorporating Reviewer feedback, enhancing dialogue diversity and difficulty. We construct a multi-turn dataset using the Alpaca dataset and fine-tune the LLaMA2-13B model. Evaluations on MT-Bench, MMLU-Pro, and Auto-Arena demonstrate significant improvements, achieving absolute gains of 2.9\% on MMLU-Pro and 2\% on MT-Bench compared to prior state-of-the-art models based on LLaMA2-13B. Ablation studies confirm the critical role of the Review stage and the use of multiple Reviewers in boosting instruction diversity and difficulty. Our work highlights the potential of review-driven, multi-agent frameworks for generating high-quality conversational data at scale.

cs.CL↗

arXiv2025

Three-dimensional Transport-induced Chemistry on Temperate sub-Neptune K2-18b, Part I: the Effects of Atmospheric Dynamics

The low equilibrium temperatures of temperate sub-Neptunes lead to extremely long chemical timescales in their upper atmospheres, causing the abundances of chemical species to be strongly shaped by atmospheric transport. Here, we used a three-dimensional (3D) general circulation model involving a passive tracer to investigate the atmospheric circulation and 3D transport of temperate gas-rich sub-Neptunes, using K2-18b as an example. We model K2-18b as a synchronous or asynchronous rotator, exploring spin-orbit resonances (SOR) of 2:1, 6:1, and 10:1. We find that the strong absorption of CO$_2$ and CH$_4$ induces a detached convective zone between 1 and 5 bar, resulting in strong vertical mixing at these levels. The upper atmosphere is dominated by eastward winds (an equatorial superrotating jet present in all simulations), leading to warmer evening terminators and approximately 20% higher passive tracer mass mixing ratios compared to the morning terminators. Rotation rates have minimal impact on the strength of global mean vertical mixing, but significantly influence the latitudinal distribution of passive tracers. For synchronous, 2:1 SOR, and 6:1 SOR simulations, passive tracers are more abundant in the upwelling branches at latitudes within 60$^\circ$, while for the 10:1 SOR simulation, strong transient eddies at high latitudes (>70$^\circ$) between 0.1 to 1 bar can transport passive tracers upward from the deep atmosphere, making them more abundant there, despite their alignment with the downwelling branch of the large-scale circulation. This study focuses on the atmospheric dynamics and its influence on passive tracer transport, while a follow-up paper will incorporate active chemical species.

astro-ph.EP↗

arXiv2025

PSELDNets: Pre-trained Neural Networks on a Large-scale Synthetic Dataset for Sound Event Localization and Detection

Sound event localization and detection (SELD) has seen substantial advancements through learning-based methods. These systems, typically trained from scratch on specific datasets, have shown considerable generalization capabilities. Recently, deep neural networks trained on large-scale datasets have achieved remarkable success in the sound event classification (SEC) field, prompting an open question of whether these advances can be extended to the development of SELD foundation models. In this paper, leveraging the power of pre-trained SEC models, we propose pre-trained SELD networks (PSELDNets) on a large-scale synthetic dataset. The synthetic dataset, generated by convolving sound events with simulated spatial room impulse responses (SRIRs), contains 1,167 hours of audio clips with an ontology of 170 sound classes. These PSELDNets are applied to various SELD scenarios. When we adapt PSELDNets to specific scenarios, particularly in cases of low-resource data, we introduce a data-efficient fine-tuning method, AdapterBit. PSELDNets are evaluated on synthetic-test-set using collected SRIRs from the TAU Spatial Room Impulse Response Database (TAU-SRIR DB) and achieve satisfactory performance. We also carried out experiments to validate the transferability of PSELDNets to three publicly available datasets and our own real-world recordings. The results demonstrate that PSELDNets surpass state-of-the-art systems across all publicly available datasets. Given the need for direction-of-arrival estimation, SELD generally relies on sufficient multi-channel audio clips. However, incorporating the AdapterBit, PSELDNets show more efficient adaptability to various scenarios using minimal multi-channel or even just monophonic audio clips, outperforming traditional fine-tuning approaches.

eess.AS↗

arXiv2025

DSAC: Distributional Soft Actor-Critic for Risk-Sensitive Reinforcement Learning

We present Distributional Soft Actor-Critic (DSAC), a distributional reinforcement learning (RL) algorithm that combines the strengths of distributional information of accumulated rewards and entropy-driven exploration from Soft Actor-Critic (SAC) algorithm. DSAC models the randomness in both action and rewards, surpassing baseline performances on various continuous control tasks. Unlike standard approaches that solely maximize expected rewards, we propose a unified framework for risk-sensitive learning, one that optimizes the risk-related objective while balancing entropy to encourage exploration. Extensive experiments demonstrate DSAC's effectiveness in enhancing agent performances for both risk-neutral and risk-sensitive control tasks.

cs.LG↗

arXiv2025

Multiwavelength Analysis of GRB 250101A: From Gamma-ray Prompt Emission to Optical Afterglow

The interaction between the relativistic jet and the circumburst medium produces a multiwavelength afterglow of a gamma-ray burst (GRBs). In this work, we present multiwavelength properties of GRB~250101A based on the observations of Swift, Fermi and Mephisto. The spectral analysis of Swift/BAT and Fermi/GBM reveals a soft prompt spectrum with a low-energy photon index of $-1.18$ and a peak energy of 33 keV, and the isotropic energy is $1.4\times10^{52}~{\rm erg}$. The prompt emission of GRB 250101A aligns with Type II GRBs in the Amati relation. Meanwhile, our analysis indicates that GRB 250101A is an X-ray-rich or X-ray-dominated GRB, with intrinsic properties suggesting that it is relatively softer than most classical GRBs. Optical observation with Mephisto, beginning 197 s post-trigger, shows a single power-law decay in $uvgriz$ bands, with $F_{ν,\mathrm{obs}} \propto t^{-0.76} ν^{-1.21}$. The observed spectral index significantly exceeds theoretical predictions under standard afterglow models, suggesting a color excess of $\sim0.216$ mag. However, combining X-ray and optical afterglow, we find that GRB 250101A is more likely a ``normal burst'' rather than an ``optical-dark burst'', and the dust extinction effect plays an important role in the optical blue bands. Furthermore, there is a structural change at $T_0+2924$ s in the optical light curve, indicating a density drop of $\sim50$ \% in the interstellar medium at a distance of $\sim0.13~{\rm pc}$. Our analysis shows that this GRB clearly shows some unique characteristics in its observed X-ray rich prompt emission as well as the circumburst environment, implying a special progenitor.

astro-ph.HE↗

arXiv2025

A Large Language Model-based Multi-Agent Framework for Analog Circuits' Sizing Relationships Extraction

In the design process of the analog circuit pre-layout phase, device sizing is an important step in determining whether an analog circuit can meet the required performance metrics. Many existing techniques extract the circuit sizing task as a mathematical optimization problem to solve and continuously improve the optimization efficiency from a mathematical perspective. But they ignore the automatic introduction of prior knowledge, fail to achieve effective pruning of the search space, which thereby leads to a considerable compression margin remaining in the search space. To alleviate this problem, we propose a large language model (LLM)-based multi-agent framework for analog circuits' sizing relationships extraction from academic papers. The search space in the sizing process can be effectively pruned based on the sizing relationship extracted by this framework. Eventually, we conducted tests on 3 types of circuits, and the optimization efficiency was improved by $2.32 \sim 26.6 \times$. This work demonstrates that the LLM can effectively prune the search space for analog circuit sizing, providing a new solution for the combination of LLMs and conventional analog circuit design automation methods.

cs.AI↗

arXiv2025

An Ejection Event Captured by VLBI During the Outburst of Swift J1727.8$-$1613

We observed a newly-discovered Galactic black hole X-ray binary Swift J1727.8$-$1613 with the European Very Long Baseline Interferometry Network (EVN) at 5 GHz. The observation was conducted immediately following a radio quenching event detected by the Karl G. Jansky Very Large Array (VLA). The visibility amplitude evolution over time reveals a large-amplitude radio flare and is consistent with an ejection event. The data can be interpreted either as a stationary component (i.e., the radio core) and a moving blob, or as two blobs moving away from the core symmetrically in opposite directions. The initial angular separation speed of the two components was estimated to 30 mas d^{-1}. We respectively fitted a single circular Gaussian model component to each of 14 sliced visibility datasets. For the case of including only European baselines, during the final hour of the EVN observation, the fitted sizes exhibited linear expansion, indicating that the measured sizes were dominated by the angular separation of the two components. The 6-h EVN observation took place in a rising phase of an even larger 4-day-long radio flare, implying that the ejection events were quite frequent and therefore continuous radio monitoring is necessary to correctly estimate the power of the transient jet. Combined with X-ray monitoring data, the radio quenching and subsequent flares/ejections were likely driven by instabilities in the inner hot accretion disk.

astro-ph.HE↗

arXiv2025

Accelerated Bayesian optimization in deep cooling atoms

Laser cooling, which cools atomic and molecular gases to near absolute zero, is the crucial initial step for nearly all atomic gas experiments. However, fast achievement of numerous sub-$μ$K cold atoms is challenging. To resolve the issue, we propose and experimentally validate an intelligent polarization gradient cooling approach enhanced by optical lattice, utilizing Maximum Hypersphere Compensation Sampling Bayesian Optimization (MHCS-BO). MHCS-BO demonstrates a twofold increase in optimization efficiency and superior prediction accuracy compared to conventional Bayesian optimization. Finally, approximate $10^8$ cold atoms at a temperature of 0.4$\pm$0.2 $μ$K can be achieved given the optimal parameters within 15 minutes. Our work provides an intelligent protocol, which can be generalized to other high-dimension parameter optimization problems, and paves way for preparation of ultracold atom in quantum experiments.

physics.atom-ph↗

arXiv2025

Wasserstein and Convex Gaussian Approximations for Non-stationary Time Series of Diverging Dimensionality

In high-dimensional time series analysis, Gaussian approximation (GA) schemes under various distance measures or on various collections of subsets of the Euclidean space play a fundamental role in a wide range of statistical inference problems. To date, most GA results for high-dimensional time series are established on hyper-rectangles and their equivalence. In this paper, by considering the 2-Wasserstein distance and the collection of all convex sets, we establish a general GA theory for a broad class of high-dimensional non-stationary (HDNS) time series, extending the scope of problems that can be addressed in HDNS time series analysis. For HDNS time series of sufficiently weak dependence and light tail, the GA rates established in this paper are either nearly optimal with respect to the dimensionality and time series length, or they are nearly identical to the corresponding best-known GA rates established for independent data. A multiplier bootstrap procedure is utilized and theoretically justified to implement our GA theory. We demonstrate by two previously undiscussed time series applications the use of the GA theory and the bootstrap procedure as unified tools for a wide range of statistical inference problems in HDNS time series analysis.

math.ST↗

arXiv2025

Convergence Analysis of the Self-Adaptive Projection Method for Variational Inequalities with Non-Lipschitz Continuous Operators

In this paper, we employ Tseng's extragradient method with the self-adaptive stepsize to solve variational inequality problems involving non-Lipschitz continuous and quasimonotone operators in real Hilbert spaces. The convergence of the proposed method is analyzed under some mild assumptions. The key advantages of the method are that it does not require the operator associated with the variational inequality to be Lipschitz continuous and that it adopts the self-adaptive stepsize. Numerical experiments are also provided to illustrate the effectiveness and superiority of the method.

math.OC↗