arXiv subjects

Lin Long

Publications and source records attributed to Lin Long.

14 recordsLinked to original sources

arXiv2026

Empirical Near-Infrared Spectral Templates from SPHEREx: Disentangling AGN and Host-Galaxy Emission

SPHEREx conducts the first all-sky near-infrared spectral survey over 0.75-5.0 $\mu$m, spanning the transition from host-galaxy starlight to AGN-heated torus emission and enabling spectroscopic identification of active galactic nuclei (AGNs). We develop a data-driven framework combining weighted non-negative matrix factorization with forward modeling to learn empirical rest-frame templates from 2,777 normal galaxies and 1,064 AGNs in the North Ecliptic Pole (NEP) field. The resulting 14-component dictionary contains seven galaxy templates spanning stellar-population continua and nebular emission and seven AGN templates capturing emission lines and warm and hot torus continua, providing an additive, operational decomposition. Inferred near-infrared host fractions correlate with independent SDSS spectral and HSC imaging decompositions (at Pearson $r=0.68$ and $r=0.59$), comparable to the SDSS-HSC correlation ($r=0.66$) and supporting the reliability of the AGN-host separation. In the NEP deep field, the same templates recover redshifts with $|z_{\rm fit}-z_{\rm spec}| < 0.01$ for 93.0% of galaxies and 93.5% of AGNs and distinguish galaxies, narrow-line AGNs, and broad-line AGNs with 81.3% accuracy. At all-sky depth, classification retains 72.9% accuracy. The near-infrared coefficients also flag questionable DESI spectral classifications and redshifts, showing how SPHEREx complements optical spectroscopy. Learned directly from survey spectrophotometry and extending beyond the $K$ band, this empirical dictionary fills a long-standing gap in near-to-mid-infrared templates and provides a data-driven foundation for population-scale studies of AGN activity and its connection to host galaxies across the SPHEREx all-sky survey.

astro-ph.GA↗

arXiv2026

dRAE: Representation Autoencoder with Hyper-Spherical Codes

In this work, we aim to discretize the high-dimensional visual representations to bridge the gap with language models - a non-trivial challenge, as existing quantization methods suffer from codebook collapse, failing to scale while preserving semantic coherence. We identify the root cause as metric mismatch: standard Euclidean codebook objectives are fundamentally misaligned with the anisotropic geometry of representation space, leading to codebook embeddings with high-variance magnitude scales and uneven angular distributions that hinder scalability. To address this, we propose Hyper-Spherical Quantization (HSQ), which decouples semantic content from feature magnitude via angular routing, preventing code assignment from being dominated by scale rather than meaning. The resulting discrete Representation Autoencoder (dRAE) achieves high-fidelity reconstruction while preserving semantic integrity and supporting scalable codebook budget. Extensive experiments demonstrate consistent performance gains as the vocabulary size scales to 131{,}072, along with 100\% codebook utilization, simplified training pipeline, and strong performance across understanding and generation tasks.

cs.CV↗

arXiv2026

WorldMemArena: Evaluating Multimodal Agent Memory Through Action-World Interaction

Multimodal large language models are increasingly deployed as long-horizon agents, where memory must do more than recall: it must track an evolving world, revise what has gone stale, and surface the right evidence at decision time. Existing benchmarks measure recall over static dialogue, collapse memory into a single end-of-task accuracy, and reduce visual observations to captions, leaving us unable to localize failures to writing, maintenance, retrieval, or use. The rise of agent harnesses that author their own memory sharpens this gap, since we have no principled way to compare hand-designed pipelines with self-managing alternatives. To close these gaps, we formulate multimodal agent memory as an Action-World Interaction Loop with an observable four-stage lifecycle, and instantiate it in WorldMemArena: 400 multi-session multimodal tasks spanning Lifelong Evolution (evolving personal and task states) and Agentic Execution (memory from real observations, actions, and feedback), annotated with gold memory points, updates, distractors, and evidence chains for stage-level diagnosis. This enables the first head-to-head comparison of long-context, manually designed (RAG and external memory systems), and harness-based memory agents. Results show that: (1) better memory writing and storage do not guarantee better performance; (2) multimodal memory still struggles to fully use visual evidence; (3) systems are unstable across domains and degrade on realistic agentic trajectories; and (4) harness memory is more flexible but remains costly and less reliable.

cs.CV↗

arXiv2025

Mock Observations for the CSST Mission: Integral Field Spectrograph--Instrument Simulation

The Chinese Space Station Survey Telescope (CSST) is a next-generation Stage-IV facility renowned for its wide field of view, high image quality, and multi-band observational capabilities. Among the five instruments onboard the CSST, the Integral Field Spectrograph (IFS) offers the unique ability to simultaneously capture spatial and spectral information across a field of view of no less than $6^{''}\times6^{''}$. Key advantages of the IFS include a high spatial resolution of $0.2^{''}$ and a broad spectral coverage from 350 to 1000 nm, making it an ideal instrument for studying physical processes in the vicinity of supermassive black holes within galaxies. To more accurately assess the technical and scientific performance of the CSST-IFS, it is essential to develop a simulation tool that incorporates realistic effects from all optical components. Such a simulation will form an integral part of the CSST-IFS data and pipeline system, enabling the development of the data reduction pipeline well ahead of actual observations. This paper presents an end-to-end simulation workflow for the CSST-IFS, incorporating a wide range of instrumental effects that may influence its spectral and imaging performance. The simulation accounts for optical diffraction effects introduced by all components, such as image slicers and slit array, as well as sub-pixel effects from gratings. It also includes various detector noises, frame-shifting effects, and charge-transfer inefficiency. Real observational conditions--such as target Doppler shift, cosmic rays, and other in-orbit operational effects--are also considered. We describe the technical implementation of the simulation and present results that quantitatively characterize key instrument parameters.

astro-ph.IM↗

arXiv2025

Joint Analysis of Optical, Near-Infrared And Mid-Infrared Variability of 4 Quasars at Redshift < 1

Amid rapid advances in time-domain astronomy, multi-wavelength (e.g., optical and infrared) time-domain studies of quasars remain scarce. Here we present a systematic analysis of four quasars initially selected by their Ks-band variability amplitudes in the VISTA Variables in the V\'{\i}a L\'actea Survey (VVV/VVVX). For these objects, we obtain complementary optical light curves from Pan-STARRS1 (PS1) and the Zwicky Transient Facility (ZTF), and W1-band light curves from the Wide-field Infrared Survey Explorer (WISE). We perform correlation analysis to study the time lags between different bands, which may be directly related to the size of the dust torus. After correcting for infrared flux contamination from the accretion disk and accounting for the redshift effect, we measure the Ks-optical and W1-optical lags for the targets VVV J1834-2925 and VVV J1845-2426. Using typical sublimation temperatures and reverberation time lags, we obtain a graphite-to-silicate grain size ratio of $\frac{a_C}{a_S}\sim$ 0.4. Through SED fitting, we determine the luminosities of these quasars and find that their dust torus sizes follow the established $R_{dust}-L_{AGN}$ relation reported in previous studies.

astro-ph.GA↗

arXiv2025

Optimizing Long-term Variability of AGN Light Curves. I. A Case Study with ZTF Observations in the EGS Field

Optical variability is a key observational probe for studying the accretion dynamics and central engine physics of Active Galactic Nuclei (AGNs). The quality and completeness of light curves have a direct impact on variability studies, particularly for faint AGNs and high-redshift AGNs. To improve the quality of long-term light curves for AGNs, we bin and stack multi-epoch images balancing the image depths and temporal resolution. As a case study, we apply this method to Zwicky Transient Facility (ZTF) observations in the Extended Groth Strip (EGS) field, where the overlapping region covers an area of about 370 arcmin$^2$ and includes $g$-band and $r$-band data taken from March 2018 to December 2024. The co-added images are approximately 2.0 to 2.5 magnitudes deeper than the ZTF single-epoch images. With co-added images, we construct light curves for 73 AGNs in the EGS field. Compared to the traditional ZTF light curves, our light curves maintain consistent long-term variability trends but with higher photometric precision. Furthermore, this method can help detect AGNs with weak variability which are missed from the traditional ZTF data due to the noisy light curves or below the detection limit in ZTF's single-epoch exposure. Among the 73 AGNs, the majority exhibit a bluer-when-brighter (BWB) trend on long-term timescales, which is consistent with previous studies. This work offers insights for optimizing AGN light curves in both current and upcoming all-sky time-domain surveys.

astro-ph.GA↗

arXiv2025

SAG-Agent: Enabling Long-Horizon Reasoning in Strategy Games via Dynamic Knowledge Graphs

Most commodity software lacks accessible Application Programming Interfaces (APIs), requiring autonomous agents to interact solely through pixel-based Graphical User Interfaces (GUIs). In this API-free setting, large language model (LLM)-based agents face severe efficiency bottlenecks: limited to local visual experiences, they make myopic decisions and rely on inefficient trial-and-error, hindering both skill acquisition and long-horizon planning. To overcome these limitations, we propose SAG-Agent, an experience-driven learning framework that structures an agent's raw pixel-level interactions into a persistent State-Action Graph (SAG). SAG-Agent mitigates inefficient exploration by topologically linking functionally similar but visually distinct GUI states, constructing a rich neighborhood of experience that enables the agent to generalize from a diverse set of historical strategies. To facilitate long-horizon reasoning, we design a novel hybrid intrinsic reward mechanism based on the graph topology, combining a state-value reward for exploiting known high-value pathways with a novelty reward that encourages targeted exploration. This approach decouples strategic planning from pure discovery, allowing the agent to effectively value setup actions with delayed gratification. We evaluate SAG-Agent in two complex, open-ended GUI-based decision-making environments (Civilization V and Slay the Spire), demonstrating significant improvements in exploration efficiency and strategic depth over the state-of-the-art methods.

cs.AI↗

arXiv2025

Understanding Language Prior of LVLMs by Contrasting Chain-of-Embedding

Large vision-language models (LVLMs) achieve strong performance on multimodal tasks, yet they often default to their language prior (LP) -- memorized textual patterns from pre-training while under-utilizing visual evidence. Prior analyses of LP mostly rely on input-output probing, which fails to reveal the internal mechanisms governing when and how vision influences model behavior. To address this gap, we present the first systematic analysis of language prior through the lens of chain-of-embedding, which examines the layer-wise representation dynamics within LVLMs. Our analysis reveals a universal phenomenon: each model exhibits a Visual Integration Point (VIP), a critical layer at which visual information begins to meaningfully reshape hidden representations and influence decoding for multimodal reasoning. Building on this observation, we introduce the Total Visual Integration (TVI) estimator, which aggregates representational discrepancy beyond the VIP to quantify how strongly visual query influences response generation. Across 60 model-dataset combinations spanning 10 contemporary LVLMs and 6 benchmarks, we demonstrate that VIP consistently emerges, and that TVI reliably predicts the strength of language prior. This offers a principled toolkit for diagnosing and understanding language prior in LVLMs.

cs.LG↗

arXiv2025

MCI: Multi-Channel Imager on the Chinese Space Station Survey Telescope

The Multi-Channel Imager (MCI) is a powerful near-ultraviolet (NUV) and visible imager onboard the Chinese Space Station Survey Telescope (CSST). The MCI provides three imaging channels, which are the NUV channel, the Blue channel and the Red channel, with the wavelength range of 255-430 nm, 430-700 nm, and 700-1000 nm, respectively. MCI's three channels can target the same field simultaneously, which is unique compared to other imagers onboard the Hubble Space Telescope (HST) or the James Webb Space Telescope (JWST). Each channel employs a CCD focal plane of 9216 x 9232 pixels and $\sim$7\arcmin.5 x 7\arcmin.5 field of view (FOV), which are about $\gtrsim 4$ times greater than the FOVs of HST imagers. The MCI's three channels feature unprecedented sensitivities and field of views complement the NUV and visible capabilities of the CSST for high-precision photometry and weak-signal detection, which would help build a new standard-star system and the deepest UV-Optical exposures for CSST. Rich filter sets of MCI would help explore other sciences such as local emission line mapping, high-z Ly$\alpha$ emitters searching, etc. Here we present key design features, results of current ground tests, and suggested observing strategies of the MCI.

astro-ph.IM↗

arXiv2025

Seeing, Listening, Remembering, and Reasoning: A Multimodal Agent with Long-Term Memory

We introduce M3-Agent, a novel multimodal agent framework equipped with long-term memory. Like humans, M3-Agent can process real-time visual and auditory inputs to build and update episodic and semantic memories, gradually accumulating world knowledge. Its memory is organized in an entity-centric, multimodal manner, enabling deeper and more consistent understanding of the environment. Given an instruction, M3-Agent autonomously performs multi-turn reasoning and retrieves relevant memories to complete tasks. To evaluate memory effectiveness and memory-based reasoning in multimodal agents, we develop M3-Bench, a long-video question answering benchmark comprising 100 newly recorded robot-perspective videos (M3-Bench-robot) and 920 diverse web-sourced videos (M3-Bench-web). We annotate QA pairs designed to test capabilities essential for agent applications, such as person understanding, general knowledge extraction, and cross-modal reasoning. Experimental results show that M3-Agent, trained via reinforcement learning, outperforms the strongest baseline, a prompting agent using Gemini-1.5-pro and GPT-4o, achieving 6.7%, 7.7%, and 5.3% higher accuracy on M3-Bench-robot, M3-Bench-web and VideoMME-long, respectively. Our work advances multimodal agents toward more human-like long-term memory and provides insights for their practical design. Model, code and data are available at https://github.com/bytedance-seed/m3-agent.

cs.CV↗

arXiv2024

TableGPT2: A Large Multimodal Model with Tabular Data Integration

The emergence of models like GPTs, Claude, LLaMA, and Qwen has reshaped AI applications, presenting vast new opportunities across industries. Yet, the integration of tabular data remains notably underdeveloped, despite its foundational role in numerous real-world domains. This gap is critical for three main reasons. First, database or data warehouse data integration is essential for advanced applications; second, the vast and largely untapped resource of tabular data offers immense potential for analysis; and third, the business intelligence domain specifically demands adaptable, precise solutions that many current LLMs may struggle to provide. In response, we introduce TableGPT2, a model rigorously pre-trained and fine-tuned with over 593.8K tables and 2.36M high-quality query-table-output tuples, a scale of table-related data unprecedented in prior research. This extensive training enables TableGPT2 to excel in table-centric tasks while maintaining strong general language and coding abilities. One of TableGPT2's key innovations is its novel table encoder, specifically designed to capture schema-level and cell-level information. This encoder strengthens the model's ability to handle ambiguous queries, missing column names, and irregular tables commonly encountered in real-world applications. Similar to visual language models, this pioneering approach integrates with the decoder to form a robust large multimodal model. We believe the results are compelling: over 23 benchmarking metrics, TableGPT2 achieves an average performance improvement of 35.20% in the 7B model and 49.32% in the 72B model over prior benchmark-neutral LLMs, with robust general-purpose capabilities intact.

cs.LG↗

arXiv2024

On LLMs-Driven Synthetic Data Generation, Curation, and Evaluation: A Survey

Within the evolving landscape of deep learning, the dilemma of data quantity and quality has been a long-standing problem. The recent advent of Large Language Models (LLMs) offers a data-centric solution to alleviate the limitations of real-world data with synthetic data generation. However, current investigations into this field lack a unified framework and mostly stay on the surface. Therefore, this paper provides an organization of relevant studies based on a generic workflow of synthetic data generation. By doing so, we highlight the gaps within existing research and outline prospective avenues for future study. This work aims to shepherd the academic and industrial communities towards deeper, more methodical inquiries into the capabilities and applications of LLMs-driven synthetic data generation.

cs.CL↗

arXiv2023

Value-added catalog of M-giant stars in LAMOST DR9

In this work, we update the catalog of M-giant stars from the low-resolution spectra of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) Data Release 9. There are 58,076 M giants identified from the classification pipeline with seven temperature subtypes from M0 to M6. The 2471 misclassified non-M-giant stars are white dwarf binaries, early types, and M dwarfs. And the contamination rate is 4.2$\%$ in the M-giants sample. A total of 372 carbon stars were identified by CaH spectral indices, and were further confirmed by the LAMOST spectra. We update the correlation between the $(W1-W2)_0$ color and [M/H] from APOGEE DR17. We calculate the radial velocities of all M giants by applying cross-correlation to the spectra between 8000 and 8950 \AA with synthetic spectra from ATLAS9. Taking star distances less than 4 kpc from Gaia EDR3 as the standard, we refitted the photometric distance relation of M giants. And based on our M-giant stars, we select a group of Sagittarius stream members, whose sky and 3D velocity distributions are well consistent with K-giant Saggitarius stream members found in Yang et al. With our M giants, we find that the disk is asymmetric out to R = 25 kpc, which is 5 kpc further out than detected using K giants.

astro-ph.SR↗

arXiv2023

Atmospheric parameters and kinematic information for the M giants stars from LAMOST DR9

A catalog of more than 43,000 M giant stars has been selected by Li et al. from the ninth data release of LAMOST. Using the data-driven method SLAM, we obtain the stellar parameters (Teff, logg, [M/H], [$\alpha$/M]) for all the M giant stars with uncertainties of 57 K, 0.25 dex, 0.16 dex and 0.06 dex at SNR > 100, respectively. With those stellar parameters, we constrain the absolute magnitude in K-band, which brings distance with relative uncertainties around 25% statistically. Radial velocities are also calculated by applying cross correlation on the spectra between 8000 A $\AA$ and 8950 A $\AA$ with synthetic spectra from ATLAS9, which covers the Ca II triplet. Comparison between our radial velocities and those from APOGEE DR17 and Gaia DR3 shows that our radial velocities have a system offset and dispersion around 1 and 4.6 km s$^{-1}$, respectively. With the distances and radial velocities combining with the astrometric data from Gaia DR3, we calculate the full 6D position and velocity information, which are able to be used for further chemo-dynamic studies on the disk and substructures in the halo, especially the Sagittarius Stream.

astro-ph.SR↗