arXiv subjects

Zihao Wu

Publications and source records attributed to Zihao Wu.

At least 73 records · Page 4Linked to original sources

arXiv2025

JADES-GS-z14-1: A Compact, Faint Galaxy at $z\approx14$ with Weak Metal Lines from Extremely Deep JWST MIRI, NIRCam, and NIRSpec Observations

JWST has shed light on galaxy formation and metal enrichment within 300 Myr of the Big Bang. While luminous galaxies at $z > 10$ often show significant [O III]$λλ$4959, 5007 emission lines, it remains unclear whether such features are prevalent among fainter, more typical galaxies due to observational limits. We present deep imaging and spectroscopy of JADES-GS-z14-1 at $z_\mathrm{spec}=13.86^{+0.04}_{-0.05}$, currently the faintest spectroscopically confirmed galaxy at $z\approx 14$. It serendipitously received 70.7 hours of MIRI/F770W imaging in the JWST Advanced Deep Extragalactic Survey (JADES), the deepest MIRI exposure for any high-redshift galaxy to date. Nonetheless, we detect only tentative F770W emission of $7.9\pm2.8$ nJy at $2.8σ$ significance, constraining the total equivalent width of [O III]$λλ$4959, 5007 + H$β$ to $520^{+400}_{-380}$ A, weaker than most $z > 10$ galaxies with MIRI detections. This source is unresolved across 16 NIRCam bands, implying a physical radius $\lesssim50$ pc. NIRSpec/PRISM spectroscopy totaling 56 hours reveals no rest-frame ultraviolet emission lines above $3 σ$. Stellar population synthesis suggests a stellar mass $\sim4\times 10^{7}$ $\mathrm{M_\odot}$ and a star formation rate $\sim 2$ $\mathrm{M_\odot yr^{-1}}$. The absence of strong metal emission lines despite intense star formation suggests a gas-phase metallicity below 10% solar and potentially a high escape fraction of ionizing photons. These deep observations provide rare constraints on faint, early galaxies, tracing the onset of chemical enrichment and ionization in the early Universe.

astro-ph.GA↗

arXiv2025

Singularity-Free Feynman Integral Bases

Standard integration-by-parts (IBP) reduction methods typically yield Feynman integral bases where the reduction of some integrals gives rise to coefficients singular as the dimensional regulator $ε\rightarrow 0$. These singular coefficients can also appear in scattering amplitudes, obscuring their structure, and rendering their evaluation more complicated. We investigate the use of bases in which the reduction of any integral is free of singular coefficients. We present two general algorithms for constructing such bases. The first is based on sequential $D=4$ IBP reduction. It constructs a basis iteratively by projecting onto the finite part of the set of IBP relations. The second algorithm performs Gaussian elimination within a local ring forbidding division by $ε$ while permitting division by polynomials in $ε$ finite at $ε=0$. We study the application of both algorithms to a pair of two-loop examples, the planar and nonplanar double-box families of integrals. We also explore the incorporation of finite Feynman integrals into these bases. In one example, the resulting basis provides a simpler and more compact representation of a scattering amplitude.

hep-th↗

arXiv2025

Argus: Leveraging Multiview Images for Improved 3-D Scene Understanding With Large Language Models

Advancements in foundation models have made it possible to conduct applications in various downstream tasks. Especially, the new era has witnessed a remarkable capability to extend Large Language Models (LLMs) for tackling tasks of 3D scene understanding. Current methods rely heavily on 3D point clouds, but the 3D point cloud reconstruction of an indoor scene often results in information loss. Some textureless planes or repetitive patterns are prone to omission and manifest as voids within the reconstructed 3D point clouds. Besides, objects with complex structures tend to introduce distortion of details caused by misalignments between the captured images and the dense reconstructed point clouds. 2D multi-view images present visual consistency with 3D point clouds and provide more detailed representations of scene components, which can naturally compensate for these deficiencies. Based on these insights, we propose Argus, a novel 3D multimodal framework that leverages multi-view images for enhanced 3D scene understanding with LLMs. In general, Argus can be treated as a 3D Large Multimodal Foundation Model (3D-LMM) since it takes various modalities as input(text instructions, 2D multi-view images, and 3D point clouds) and expands the capability of LLMs to tackle 3D tasks. Argus involves fusing and integrating multi-view images and camera poses into view-as-scene features, which interact with the 3D features to create comprehensive and detailed 3D-aware scene embeddings. Our approach compensates for the information loss while reconstructing 3D point clouds and helps LLMs better understand the 3D world. Extensive experiments demonstrate that our method outperforms existing 3D-LMMs in various downstream tasks.

cs.CV↗

arXiv2025

A new method for finding more symmetry relations of Feynman integrals

We introduce a new method for deriving Feynman integral symmetry relations. By solving the ansatz of momentum transformation in the field of rational functions rather than constants, this method can sometimes find more symmetry relations, compared to some state-of-the-art software. The new method may help to further decrease the number of unique sectors in an integral family. Well-chosen gauge conditions are implemented in this method for the efficient symmetry searching.

hep-ph↗

arXiv2025

Kira 3: integral reduction with efficient seeding and optimized equation selection

We present version 3 of Kira, a Feynman integral reduction program for high-precision calculations in quantum field theory and gravitational-wave physics. Building on previous versions, Kira 3 introduces optimized seeding and equation selection algorithms, significantly improving performance for multi-loop and multi-scale problems. New features include convenient numerical sampling, symbolic integration-by-parts reductions, and support for user-defined additional relations. We demonstrate its capabilities through benchmarks on two- and three-loop topologies, showcasing up to two orders of magnitude improvement over Kira 2.3. Kira 3 is publicly available and poised to support ambitious projects in particle physics and beyond.

hep-ph↗

arXiv2025

Detecting Intermediate-mass Black Holes Using Quasar Microlensing

Recent studies suggest that numerous intermediate-mass black holes (IMBHs) may wander undetected across the Universe, emitting little radiation. These IMBHs largely preserve their birth masses, offering critical insights into the formation of heavy black hole seeds and the dynamical processes driving their evolution. We propose that such IMBHs could produce detectable microlensing effects on quasars. Their Einstein radii, comparable to the scale of quasar broad-line regions, magnify radiation from the accretion disk and broad emission lines, making these quasars outliers in flux scaling relations. Meanwhile, the microlensing causes long-term, quasi-linear variability that is distinguishable from the stochastic variability of quasars through its coherent multi-wavelength behavior. We develop a matched-filtering technique that effectively separates the long-term lensing signal from the intrinsic quasar variability, with sensitivity tripling each time the observational time span doubles. Moreover, as IMBHs are often surrounded by dense star clusters, their combined gravitational field produces substantial extended, concentric caustics. These caustics induce significant variability in optical, ultraviolet, and X-ray bands over decade timescales, alongside hour-to-day-scale flux fluctuations in broad emission lines. We predict a substantial number of detectable events in the upcoming surveys by the Vera C. Rubin Observatory, considering recent IMBH mass density estimates. Even in the absence of positive detections, searches for these microlensing signals will place meaningful constraints on the cosmological mass density of IMBHs, advancing our understanding of their role in cosmic evolution.

astro-ph.GA↗

arXiv2025

Autono: A ReAct-Based Highly Robust Autonomous Agent Framework

This paper proposes a highly robust autonomous agent framework based on the ReAct paradigm, designed to solve complex tasks through adaptive decision making and multi-agent collaboration. Unlike traditional frameworks that rely on fixed workflows generated by LLM-based planners, this framework dynamically generates next actions during agent execution based on prior trajectories, thereby enhancing its robustness. To address potential termination issues caused by adaptive execution paths, I propose a timely abandonment strategy incorporating a probabilistic penalty mechanism. For multi-agent collaboration, I introduce a memory transfer mechanism that enables shared and dynamically updated memory among agents. The framework's innovative timely abandonment strategy dynamically adjusts the probability of task abandonment via probabilistic penalties, allowing developers to balance conservative and exploratory tendencies in agent execution strategies by tuning hyperparameters. This significantly improves adaptability and task execution efficiency in complex environments. Additionally, agents can be extended through external tool integration, supported by modular design and MCP protocol compatibility, which enables flexible action space expansion. Through explicit division of labor, the multi-agent collaboration mechanism enables agents to focus on specific task components, thereby significantly improving execution efficiency and quality.

cs.MA↗

arXiv2025

Bhakti: A Lightweight Vector Database Management System for Endowing Large Language Models with Semantic Search Capabilities and Memory

With the rapid development of big data and artificial intelligence technologies, the demand for effective processing and retrieval of vector data is growing. Against this backdrop, I have developed the Bhakti vector database, aiming to provide a lightweight and easy-to-deploy solution to meet the storage and semantic search needs of small and medium-sized datasets. Bhakti supports a variety of similarity calculation methods and a domain-specific language (DSL) for document-based pattern matching pre-filtering, facilitating migration of data with its portable data files, flexible data management and seamless integration with Python3. Furthermore, I propose a memory-enhanced large language model dialogue solution based on the Bhakti database, which can assign different weights to the question and answer in dialogue history, achieving fine-grained control over the semantic importance of each segment in a single dialogue history. Through experimental validation, my method shows significant performance in the application of semantic search and question-answering systems. Although there are limitations in processing large datasets, such as not supporting approximate calculation methods like HNSW, the lightweight nature of Bhakti gives it a clear advantage in scenarios involving small and medium-sized datasets.

cs.DB↗

arXiv2025

Deciphering the Nature of Virgil: An Obscured AGN Lurking Within an Apparently Normal Lyman-α Emitter During Cosmic Reionization

We present a comprehensive analysis of the MIRI Extremely Red Object Virgil, a Lyman-$α$ emitter at $z_{spec} = 6.6379 \pm 0.0035$ with the photometric properties of a Little Red Dot. Leveraging new JWST/MIRI imaging from the MIDIS and PAHSPECS programs, we confirm Virgil's extraordinary nature among galaxies in JADES/GOODS-South, exhibiting a strikingly red NIRCam-to-MIRI color (F444W $-$ F1500W = $2.84\pm0.04$ mag). Deep NIRSpec/PRISM spectroscopy from the OASIS program offers key insights into the host galaxy, revealing properties of an average star-forming galaxy during Cosmic Reionization, such as a subsolar metallicity, low-to-moderate dust content, and a relatively high ionization parameter and electron temperature. By estimating the star formation rate of Virgil from UV and H$α$, we find evidence that the galaxy is either entering or fading out of a bursty episode. Although line-ratio diagnostics employed at high-$z$ would classify Virgil as an Active Galactic Nucleus (AGN), this classification becomes ambiguous once redshift evolution is considered. Nonetheless, Virgil occupies the same parameter space as recently confirmed AGNs at similar redshifts. The new deep MIRI data at 15 $μ$m reinforce the AGN nature of Virgil, as inferred from multiple spectral energy distribution (SED) fitting codes. Virgil's rising infrared SED and UV excess resemble those of Dust-Obscured Galaxies (DOGs) studied with Spitzer at Cosmic Noon, particularly blue-excess HotDOGs. Our results highlight the need for a multi-wavelength approach incorporating MIRI to uncover such extreme sources at $z\gtrsim6$ and to shed light on the interplay between galaxy evolution and early black hole growth during Cosmic Reionization.

astro-ph.GA↗

arXiv2025

Score-Based Metropolis-Hastings Algorithms

In this paper, we introduce a new approach for integrating score-based models with the Metropolis-Hastings algorithm. While traditional score-based diffusion models excel in accurately learning the score function from data points, they lack an energy function, making the Metropolis-Hastings adjustment step inaccessible. Consequently, the unadjusted Langevin algorithm is often used for sampling using estimated score functions. The lack of an energy function then prevents the application of the Metropolis-adjusted Langevin algorithm and other Metropolis-Hastings methods, limiting the wealth of other algorithms developed that use acceptance functions. We address this limitation by introducing a new loss function based on the \emph{detailed balance condition}, allowing the estimation of the Metropolis-Hastings acceptance probabilities given a learned score function. We demonstrate the effectiveness of the proposed method for various scenarios, including sampling from heavy-tail distributions.

cs.LG↗

arXiv2025

Full-colour double-virtual amplitudes for associated production of a Higgs boson with a bottom-quark pair at the LHC

We present the double-virtual amplitudes contributing to the production of a Higgs boson in association with a $b\bar{b}$ pair at the Large Hadron Collider. We perform the computation within the five-flavour scheme, which employs massless bottom quarks and finite bottom-Yukawa coupling, taking into account all the colour structures. We derive the analytic form of the helicity amplitudes through finite-field reconstruction techniques. The analytic expressions have been implemented in a public C++ library, and we demonstrate that evaluations are sufficiently stable and efficient for use in phenomenological studies.

hep-ph↗

arXiv2025

Performing integration-by-parts reductions using NeatIBP 1.1 + Kira

We introduce a new version v1.1 of NeatIBP. In this new version, a Kira interface is included. It allows the user to reduce the integration-by-parts (IBP) identity systems generated by NeatIBP using Kira in a highly automated way. This new version also implements the so-called spanning cuts method. It helps to reduce the total computational complexity of IBP reduction for certain hard problems. Another important feature of this new version is an algorithm to simplify the solution module of the syzygy equations hinted by the idea of maximal cuts.

hep-ph↗

arXiv2025

S2TX: Cross-Attention Multi-Scale State-Space Transformer for Time Series Forecasting

Time series forecasting has recently achieved significant progress with multi-scale models to address the heterogeneity between long and short range patterns. Despite their state-of-the-art performance, we identify two potential areas for improvement. First, the variates of the multivariate time series are processed independently. Moreover, the multi-scale (long and short range) representations are learned separately by two independent models without communication. In light of these concerns, we propose State Space Transformer with cross-attention (S2TX). S2TX employs a cross-attention mechanism to integrate a Mamba model for extracting long-range cross-variate context and a Transformer model with local window attention to capture short-range representations. By cross-attending to the global context, the Transformer model further facilitates variate-level interactions as well as local/global communications. Comprehensive experiments on seven classic long-short range time-series forecasting benchmark datasets demonstrate that S2TX can achieve highly robust SOTA results while maintaining a low memory footprint.

cs.LG↗

arXiv2025

Teleportation With Null Space Gradient Projection for Optimization Acceleration

Optimization techniques have become increasingly critical due to the ever-growing model complexity and data scale. In particular, teleportation has emerged as a promising approach, which accelerates convergence of gradient descent-based methods by navigating within the loss invariant level set to identify parameters with advantageous geometric properties. Existing teleportation algorithms have primarily demonstrated their effectiveness in optimizing Multi-Layer Perceptrons (MLPs), but their extension to more advanced architectures, such as Convolutional Neural Networks (CNNs) and Transformers, remains challenging. Moreover, they often impose significant computational demands, limiting their applicability to complex architectures. To this end, we introduce an algorithm that projects the gradient of the teleportation objective function onto the input null space, effectively preserving the teleportation within the loss invariant level set and reducing computational cost. Our approach is readily generalizable from MLPs to CNNs, transformers, and potentially other advanced architectures. We validate the effectiveness of our algorithm across various benchmark datasets and optimizers, demonstrating its broad applicability.

cs.LG↗

arXiv2025

Photometric detection at $7.7\ μ\mathrm{m}$ of a galaxy beyond redshift $14$ with JWST/MIRI

The James Webb Space Telescope (JWST) has spectroscopically confirmed numerous galaxies at $z > 10$. While weak rest-ultraviolet emission lines have only been seen in a handful of sources, the stronger rest-optical emission lines are highly diagnostic and accessible at mid-infrared wavelengths with the Mid-Infrared Instrument (MIRI) of JWST. We report the photometric detection of the distant spectroscopically confirmed galaxy JADES-GS-z14-0 at $z = 14.32^{+0.08}_{-0.20}$ with MIRI at $7.7\ μ\mathrm{m}$. The most plausible solution for the stellar population properties is that this galaxy contains half a billion solar masses in stars with a strong burst of star formation in the most recent few million years. For this model, at least one-third of the flux at $7.7\ μ\mathrm{m}$ comes from the rest-optical emission lines $\mathrm{H}β$ and/or $\mathrm{[OIII]}λ\lambda4959,5007$. The inferred properties of JADES-GS-z14-0 suggest rapid mass assembly and metal enrichment during the earliest phases of galaxy formation. This work demonstrates the unique power of mid-infrared observations in understanding galaxies at the redshift frontier.

astro-ph.GA↗

arXiv2025

Large Language Models for Bioinformatics

With the rapid advancements in large language model (LLM) technology and the emergence of bioinformatics-specific language models (BioLMs), there is a growing need for a comprehensive analysis of the current landscape, computational characteristics, and diverse applications. This survey aims to address this need by providing a thorough review of BioLMs, focusing on their evolution, classification, and distinguishing features, alongside a detailed examination of training methodologies, datasets, and evaluation frameworks. We explore the wide-ranging applications of BioLMs in critical areas such as disease diagnosis, drug discovery, and vaccine development, highlighting their impact and transformative potential in bioinformatics. We identify key challenges and limitations inherent in BioLMs, including data privacy and security concerns, interpretability issues, biases in training data and model outputs, and domain adaptation complexities. Finally, we highlight emerging trends and future directions, offering valuable insights to guide researchers and clinicians toward advancing BioLMs for increasingly sophisticated biological and clinical applications.

q-bio.QM↗

arXiv2024

MGH Radiology Llama: A Llama 3 70B Model for Radiology

In recent years, the field of radiology has increasingly harnessed the power of artificial intelligence (AI) to enhance diagnostic accuracy, streamline workflows, and improve patient care. Large language models (LLMs) have emerged as particularly promising tools, offering significant potential in assisting radiologists with report generation, clinical decision support, and patient communication. This paper presents an advanced radiology-focused large language model: MGH Radiology Llama. It is developed using the Llama 3 70B model, building upon previous domain-specific models like Radiology-GPT and Radiology-Llama2. Leveraging a unique and comprehensive dataset from Massachusetts General Hospital, comprising over 6.5 million de-identified medical reports across various imaging modalities, the model demonstrates significant improvements in generating accurate and clinically relevant radiology impressions given the corresponding findings. Our evaluation, incorporating both traditional metrics and a GPT-4-based assessment, highlights the enhanced performance of this work over general-purpose LLMs.

cs.CL↗

arXiv2024

Offline Stochastic Optimization of Black-Box Objective Functions

Many challenges in science and engineering, such as drug discovery and communication network design, involve optimizing complex and expensive black-box functions across vast search spaces. Thus, it is essential to leverage existing data to avoid costly active queries of these black-box functions. To this end, while Offline Black-Box Optimization (BBO) is effective for deterministic problems, it may fall short in capturing the stochasticity of real-world scenarios. To address this, we introduce Stochastic Offline BBO (SOBBO), which tackles both black-box objectives and uncontrolled uncertainties. We propose two solutions: for large-data regimes, a differentiable surrogate allows for gradient-based optimization, while for scarce-data regimes, we directly estimate gradients under conservative field constraints, improving robustness, convergence, and data efficiency. Numerical experiments demonstrate the effectiveness of our approach on both synthetic and real-world tasks.

cs.LG↗