Searcharxiv⌕ Search

arXiv subjects

Bo Zhang

Publications and source records attributed to Bo Zhang.

At least 181 records · Page 10Linked to original sources
arXiv2026

InternAgent-1.5: A Unified Agentic Framework for Long-Horizon Autonomous Scientific Discovery

We introduce InternAgent-1.5, a unified system designed for end-to-end scientific discovery across computational and empirical domains. The system is built on a structured architecture composed of three coordinated subsystems for generation, verification, and evolution. These subsystems are supported by foundational capabilities for deep research, solution optimization, and long horizon memory. The architecture allows InternAgent-1.5 to operate continuously across extended discovery cycles while maintaining coherent and improving behavior. It also enables the system to coordinate computational modeling and laboratory experimentation within a single unified system. We evaluate InternAgent-1.5 on scientific reasoning benchmarks such as GAIA, HLE, GPQA, and FrontierScience, and the system achieves leading performance that demonstrates strong foundational capabilities. Beyond these benchmarks, we further assess two categories of discovery tasks. In algorithm discovery tasks, InternAgent-1.5 autonomously designs competitive methods for core machine learning problems. In empirical discovery tasks, it executes complete computational or wet lab experiments and produces scientific findings in earth, life, biological, and physical domains. Overall, these results show that InternAgent-1.5 provides a general and scalable framework for autonomous scientific discovery.

cs.AI↗
arXiv2026

SciDataCopilot: An Agentic Data Preparation Framework for AGI-driven Scientific Discovery

The current landscape of AI for Science (AI4S) is predominantly anchored in large-scale textual corpora, where generative AI systems excel at hypothesis generation, literature search, and multi-modal reasoning. However, a critical bottleneck for accelerating closed-loop scientific discovery remains the utilization of raw experimental data. Characterized by extreme heterogeneity, high specificity, and deep domain expertise requirements, raw data possess neither direct semantic alignment with linguistic representations nor structural homogeneity suitable for a unified embedding space. The disconnect prevents the emerging class of Artificial General Intelligence for Science (AGI4S) from effectively interfacing with the physical reality of experimentation. In this work, we extend the text-centric AI-Ready concept to Scientific AI-Ready data paradigm, explicitly formalizing how scientific data is specified, structured, and composed within a computational workflow. To operationalize this idea, we propose SciDataCopilot, an autonomous agentic framework designed to handle data ingestion, scientific intent parsing, and multi-modal integration in a end-to-end manner. By positioning data readiness as a core operational primitive, the framework provides a principled foundation for reusable, transferable systems, enabling the transition toward experiment-driven scientific general intelligence. Extensive evaluations across three heterogeneous scientific domains show that SciDataCopilot improves efficiency, scalability, and consistency over manual pipelines, with up to 30$\times$ speedup in data preparation.

cs.DB↗
arXiv2026

Halt the Hallucination: Decoupling Signal and Semantic OOD Detection Based on Cascaded Early Rejection

Efficient and robust Out-of-Distribution (OOD) detection is paramount for safety-critical applications.However, existing methods still execute full-scale inference on low-level statistical noise. This computational mismatch not only incurs resource waste but also induces semantic hallucination, where deep networks forcefully interpret physical anomalies as high-confidence semantic features.To address this, we propose the Cascaded Early Rejection (CER) framework, which realizes hierarchical filtering for anomaly detection via a coarse-to-fine logic.CER comprises two core modules: 1)Structural Energy Sieve (SES), which establishes a non-parametric barrier at the network entry using the Laplacian operator to efficiently intercept physical signal anomalies; and 2) the Semantically-aware Hyperspherical Energy (SHE) detector, which decouples feature magnitude from direction in intermediate layers to identify fine-grained semantic deviations. Experimental results demonstrate that CER not only reduces computational overhead by 32% but also achieves a significant performance leap on the CIFAR-100 benchmark:the average FPR95 drastically decreases from 33.58% to 22.84%, and AUROC improves to 93.97%. Crucially, in real-world scenarios simulating sensor failures, CER exhibits performance far exceeding state-of-the-art methods. As a universal plugin, CER can be seamlessly integrated into various SOTA models to provide performance gains.

cs.CV↗
arXiv2026

Convergence of the PML method for thermoelastic wave scattering problems

This paper is concerned with the thermoelastic obstacle scattering problem in three dimensions. A uniaxial perfectly matched layer (PML) method is firstly introduced to truncate the unbounded scattering problem, leading to a truncated PML problem in a bounded domain. Under certain constraints on model parameters, the well-posedness for the truncated PML problem is then proved except possibly for a discrete set of frequencies, based on the analytic Fredholm theory. Moreover, the exponential convergence of the uniaxial PML method is established in terms of the thickness and absorbing parameters of PML layer. The proof is based on the PML extension technique and the exponential decay properties of the modified fundamental solution. As far as we know, this is the first convergence result of the PML method for the time-harmonic thermoelastic scattering problem.

math.AP↗
arXiv2026

ERNIE 5.0 Technical Report

In this report, we introduce ERNIE 5.0, a natively autoregressive foundation model desinged for unified multimodal understanding and generation across text, image, video, and audio. All modalities are trained from scratch under a unified next-group-of-tokens prediction objective, based on an ultra-sparse mixture-of-experts (MoE) architecture with modality-agnostic expert routing. To address practical challenges in large-scale deployment under diverse resource constraints, ERNIE 5.0 adopts a novel elastic training paradigm. Within a single pre-training run, the model learns a family of sub-models with varying depths, expert capacities, and routing sparsity, enabling flexible trade-offs among performance, model size, and inference latency in memory- or time-constrained scenarios. Moreover, we systematically address the challenges of scaling reinforcement learning to unified foundation models, thereby guaranteeing efficient and stable post-training under ultra-sparse MoE architectures and diverse multimodal settings. Extensive experiments demonstrate that ERNIE 5.0 achieves strong and balanced performance across multiple modalities. To the best of our knowledge, among publicly disclosed models, ERNIE 5.0 represents the first production-scale realization of a trillion-parameter unified autoregressive model that supports both multimodal understanding and generation. To facilitate further research, we present detailed visualizations of modality-agnostic expert routing in the unified model, alongside comprehensive empirical analysis of elastic training, aiming to offer profound insights to the community.

cs.CL↗
arXiv2026

Construction-Verification: A Benchmark for Applied Mathematics in Lean 4

Recent advances in large language models have demonstrated impressive capabilities in mathematical formalization. However, existing benchmarks focus on logical verification of declarative propositions, often neglecting the task of explicitly synthesizing solutions. This limitation is particularly acute in applied mathematics domains, where the goal is frequently to derive concrete values or executable algorithms rather than solely proving theorems. To address this, we introduce a Lean 4 framework that enforces a construction-verification workflow, compelling the agent to define explicit solutions before proving their correctness. We curate a comprehensive benchmark AMBER (Applied Mathematics BEnchmark for Reasoning) spanning core domains of applied mathematics, including convex analysis, optimization, numerical algebra, and high-dimensional probability. Aside from theorem proving, our benchmark features complex tasks such as evaluation, algorithm design, and representation transformation. Experiments reveal that current models face significant difficulties with these constructive tasks. Notably, we observe that general-purpose reasoning models consistently outperform specialized theorem provers. We attribute this to a degradation of instruction following capabilities in specialized models. Fine-tuning on proof corpora appears to induce ``tactical overfitting", compromising the ability to adhere to complex constructive requirements, whereas general models retain the versatility needed for multi-task formal reasoning.

cs.LO↗
arXiv2026

Geometry-driven splitting dynamics of a triply quantized vortex in a ring-shaped condensate

We study the splitting dynamics of a triply quantized vortex (TQV) confined in a ring-shaped Bose-Einstein condensate under a weakly elliptical harmonic trap. Using full 3D simulations in cylindrical coordinates, combined with a semi-analytical energy analysis, we show that the vortex preferentially splits along the long axis of the trap, a direction that minimizes the kinetic-energy cost relative to the initial TQV state. Systematic parameter scans reveal that initial quantum fluctuations increase the splitting time and suppress the transient three-core pattern observed in noise-free simulations, whereas stronger nonlinear interactions accelerate the splitting. When the trap is nearly isotropic, the unstable Bogoliubov modes are dominated by both azimuthal quantum number $l_q=3$ and $l_q=2$; this leads to a dynamical sequence where three daughter vortices first form a triangular arrangement, later evolving into a linear chain. For stronger anisotropy, geometric coupling selectively enhances the $l_q=2$ mode, making it the sole dominant channel and resulting directly in linear vortex alignment -- a clear signature of geometry-induced mode competition explained through combined energy-based and Bogoliubov stability analysis. Our results provide a quantitative picture of how trap geometry can steer the instability pathway, splitting time, and final pattern of a multiply quantized vortex, offering a route toward geometry-controlled vortex engineering.

cond-mat.quant-gas↗
arXiv2026

SATA: Sparsity-Aware Scheduling for Selective Token Attention

Transformers have become the foundation of numerous state-of-the-art AI models across diverse domains, thanks to their powerful attention mechanism for modeling long-range dependencies. However, the quadratic scaling complexity of attention poses significant challenges for efficient hardware implementation. While techniques such as quantization and pruning help mitigate this issue, selective token attention offers a promising alternative by narrowing the attention scope to only the most relevant tokens, reducing computation and filtering out noise. In this work, we propose SATA, a locality-centric dynamic scheduling scheme that proactively manages sparsely distributed access patterns from selective Query-Key operations. By reordering operand flow and exploiting data locality, our approach enables early fetch and retirement of intermediate Query/Key vectors, improving system utilization. We implement and evaluate our token management strategy in a control and compute system, using runtime traces from selective-attention-based models. Experimental results show that our method improves system throughput by up to 1.76x and boosts energy efficiency by 2.94x, while incurring minimal scheduling overhead.

cs.AR↗
arXiv2026

Reducing End-to-End Latency of Cause-Effect Chains with Shared Cache Analysis

Cause-effect chains, as a widely used modeling method in real-time embedded systems, are extensively applied in various safety-critical domains. End-to-end latency, as a key real-time attribute of cause-effect chains, is crucial in many applications. But the analysis of end-to-end latency for cause-effect chains on multicore platforms with shared caches still presents an unresolved issue. Traditional methods typically assume that the worst-case execution time (WCET) of each task in the cause-effect chain is known. However, in the absence of scheduling information, these methods often assume that all shared cache accesses result in misses, leading to an overestimation of WCET and, consequently, affecting the accuracy of end-to-end latency. However, effectively integrating scheduling information into the WCET analysis process of the chains may introduce two challenges: first, how to leverage the structural characteristics of the chains to optimize shared cache analysis, and second, how to improve analysis accuracy while avoiding state space explosion. To address these issues, this paper proposes a novel end-to-end latency analysis framework designed for multi-chain systems on multicore platforms with shared caches. This framework extracts scheduling information and structural characteristics of cause-effect chains, constructing fine-grained and scalable inter-core memory access contexts at the basic block level for time-sensitive shared cache analysis. This results in more accurate WCET (TSC-WCET) estimates, which are then used to derive the end-to-end latency. Finally, we conduct experiments on dual-core and quad-core systems with various cache configurations, which show that under certain settings, the average maximum end-to-end latency of cause-effect chains is reduced by up to 34% and 26%.

eess.SY↗
arXiv2026

Building Self-Evolving Agents via Experience-Driven Lifelong Learning: A Framework and Benchmark

As AI advances toward general intelligence, the focus is shifting from systems optimized for static tasks to creating open-ended agents that learn continuously. In this paper, we introduce Experience-driven Lifelong Learning (ELL), a framework for building self-evolving agents capable of continuous growth through real-world interaction. The framework is built on four core principles: (1) Experience Exploration: Agents learn through continuous, self-motivated interaction with dynamic environments, navigating interdependent tasks and generating rich experiential trajectories. (2) Long-term Memory: Agents preserve and structure historical knowledge, including personal experiences, domain expertise, and commonsense reasoning, into a persistent memory system. (3) Skill Learning: Agents autonomously improve by abstracting recurring patterns from experience into reusable skills, which are actively refined and validated for application in new tasks. (4) Knowledge Internalization: Agents internalize explicit and discrete experiences into implicit and intuitive capabilities as "second nature". We also introduce StuLife, a benchmark dataset for ELL that simulates a student's holistic college journey, from enrollment to academic and personal development, across three core phases and ten detailed sub-scenarios. StuLife is designed around three key paradigm

cs.AI↗
arXiv2026

Kirkwood-Dirac Quasiprobability as a Universal Framework for Quantum Measurements Across All Regimes

The question of when the Kirkwood-Dirac quasiprobability serves as the most appropriate description for quantum measurements has remained unresolved, particularly across different measurement strengths. While known to generate anomalous weak values in the weak measurement regime and to reduce to classical probabilities under projective measurement, the physical mechanism governing its continuous transformation has been lacking. Here we demonstrate that the KD quasiprobability provides a general framework for all measurement regimes by identifying pointer-induced decoherence as the universal mechanism controlling this transition. We show that the decoherence factor F(t) simultaneously quantifies the loss of quantum coherence and interpolates the measurement strength from weak to strong. Within this framework, the KD quasiprobability naturally deforms from its full complex form-governing weak values-to the real, non-negative Wigner formula describing projective measurements, while maintaining informational completeness throughout the transition. Our work resolves the fundamental question of the KD distribution's applicability by establishing it as the universal framework that seamlessly connects all quantum measurement regimes through a physically transparent decoherence pathway.

quant-ph↗
arXiv2026

A theoretical and computational framework for three dimensional inverse medium scattering using the linearized low-rank structure

In this work we propose a theoretical and computational framework for solving the three dimensional inverse medium scattering problem, based on a set of data-driven basis arising from the linearized problem. This set of data-driven basis consists of generalizations of prolate spheroidal wave functions to three dimensions (3D PSWFs), the main ingredients to explore a low-rank approximation of the inverse solution. We first establish the fundamentals of the inverse scattering analysis, including regularity in a customized Sobolev space and new a priori estimate. This is followed by a computational framework showcasing computing the 3D PSWFs and the low-rank approximation of the inverse solution. These results rely heavily on the fact that the 3D PSWFs are eigenfunctions of both a restricted Fourier integral operator and a Sturm-Liouville differential operator. Furthermore we propose a Tikhonov regularization method with a customized penalty norm and a localized imaging technique to image a targeting object despite the possible presence of its surroundings. Finally various numerical examples are provided to demonstrate the potential of the proposed method.

math.NA↗
arXiv2026

Analytic Model of Trans-axial Sensitivity in Cylindrical PET Systems Based on Solid Angle

In positron emission tomography (PET), a clear theoretical model describing how system sensitivity varies as a source is moved trans-axially within the field of view (FOV) is lacking. The current understanding and practical intuition often suggest that sensitivity is maximum at the center of the FOV, an assumption reflected in standardized protocols. In this work, we derive an analytic model for the trans-axial-plane sensitivity distribution in a cylindrical PET scanner based on solid angle. The model, formulated as a function of trans-axial offset from the center, is validated through both Monte Carlo simulations and physical experiments on a representative system. We find that the derived theoretical distribution is essentially consistent with simulation and experimental results, revealing a non-intuitive feature: sensitivity increases with trans-axial offset, peaks at the edge of the FOV, and drops off sharply beyond it. This study provides the first closed-form model of trans-axial geometric sensitivity in cylindrical PET scanners, offering a vital benchmark for isolating detector technology improvements and revealing a non-intuitive, offset-dependent sensitivity profile that enables new protocol optimization strategies.

physics.med-ph↗
arXiv2026

Masked Face Recognition under Different Backbones

Erratum to the paper (Zhang et al., 2025): corrections to Table IV and the data in Page 3, Section A. In the post-pandemic era, a high proportion of civil aviation passengers wear masks during security checks, posing significant challenges to traditional face recognition models. The backbone network serves as the core component of face recognition models. In standard tests, r100 series models excelled (98%+ accuracy at 0.01% FAR in face comparison, high top1/top5 in search). r50 ranked second, r34_mask_v1 lagged. In masked tests, r100_mask_v2 led (90.07% accuracy), r50_mask_v3 performed best among r50 but trailed r100. Vit-Small/Tiny showed strong masked performance with gains in effectiveness. Through extensive comparative experiments, this paper conducts a comprehensive evaluation of several core backbone networks, aiming to reveal the impacts of different models on face recognition with and without masks, and provide specific deployment recommendations.

cs.CV↗
arXiv2026

Manalyzer: End-to-end Automated Meta-analysis with Multi-agent System

Meta-analysis is a systematic research methodology that synthesizes data from multiple existing studies to derive comprehensive conclusions. This approach not only mitigates limitations inherent in individual studies but also facilitates novel discoveries through integrated data analysis. Traditional meta-analysis involves a complex multi-stage pipeline including literature retrieval, paper screening, and data extraction, which demands substantial human effort and time. However, while LLM-based methods can accelerate certain stages, they still face significant challenges, such as hallucinations in paper screening and data extraction. In this paper, we propose a multi-agent system, Manalyzer, which achieves end-to-end automated meta-analysis through tool calls. The hybrid review, hierarchical extraction, self-proving, and feedback checking strategies implemented in Manalyzer significantly alleviate these two hallucinations. To comprehensively evaluate the performance of meta-analysis, we construct a new benchmark comprising 729 papers across 3 domains, encompassing text, image, and table modalities, with over 10,000 data points. Extensive experiments demonstrate that Manalyzer achieves significant performance improvements over the LLM baseline in multi meta-analysis tasks. Project page: https://black-yt.github.io/meta-analysis-page/ .

cs.AI↗
arXiv2026

FlowSearch: Advancing deep research with dynamic structured knowledge flow

Deep research is an inherently challenging task that demands both breadth and depth of thinking. It involves navigating diverse knowledge spaces and reasoning over complex, multi-step dependencies, which presents substantial challenges for agentic systems. To address this, we propose FlowSearch, a multi-agent framework that actively constructs and evolves a dynamic structured knowledge flow to drive subtask execution and reasoning. FlowSearch is capable of strategically planning and expanding the knowledge flow to enable parallel exploration and hierarchical task decomposition, while also adjusting the knowledge flow in real time based on feedback from intermediate reasoning outcomes and insights. FlowSearch achieves competitive performance on both general and scientific benchmarks, including GAIA, HLE, GPQA and TRQA, demonstrating its effectiveness in multi-disciplinary research scenarios and its potential to advance scientific discovery. The code is available at https://github.com/InternScience/InternAgent.

cs.AI↗
arXiv2026

JAX-Shock: A Differentiable, GPU-Accelerated, Shock-Capturing Neural Solver for Compressible Flow Simulation

Understanding shock-solid interactions remains a central challenge in compressiblefluiddynamics. WepresentJAX-Shock: afully-differentiable,GPU-accelerated, high-order shock-capturing solver for efficient simulation of the compressible Navier-Stokes equations. Built entirely in JAX, the framework leverages automatic differentiation to enable gradient-based optimization, parameter inference, and end-to-end training of deep learning-augmented models. The solver integrates fifth-order WENO reconstruction with an HLLC flux to resolve shocks and discontinuities with high fidelity. To handle complex geometries, an immersed boundary method is implemented for accurate representation of solid interfaces within the compressible flow field. In addition, we introduce a neural flux module trained to augment the numerical fluxes with data-driven corrections, significantly improving accuracy and generalization. JAX-Shock also supports sequence-to-sequence learning for shock interaction prediction and reverse-mode inference to identify key physical parameters from data. Compared with purely data-driven approaches, JAX-Shock enhances generalization while preserving physical consistency. The framework establishes a flexible platform for differentiable physics, learning-based modeling, and inverse design in compressible flow regimes dominated by complex shock-solid interactions.

physics.flu-dyn↗
arXiv2026

SciEvalKit: An Open-source Evaluation Toolkit for Scientific General Intelligence

We introduce SciEvalKit, a unified benchmarking toolkit designed to evaluate AI models for science across a broad range of scientific disciplines and task capabilities. Unlike general-purpose evaluation platforms, SciEvalKit focuses on the core competencies of scientific intelligence, including Scientific Multimodal Perception, Scientific Multimodal Reasoning, Scientific Multimodal Understanding, Scientific Symbolic Reasoning, Scientific Code Generation, Science Hypothesis Generation and Scientific Knowledge Understanding. It supports six major scientific domains, spanning from physics and chemistry to astronomy and materials science. SciEvalKit builds a foundation of expert-grade scientific benchmarks, curated from real-world, domain-specific datasets, ensuring that tasks reflect authentic scientific challenges. The toolkit features a flexible, extensible evaluation pipeline that enables batch evaluation across models and datasets, supports custom model and dataset integration, and provides transparent, reproducible, and comparable results. By bridging capability-based evaluation and disciplinary diversity, SciEvalKit offers a standardized yet customizable infrastructure to benchmark the next generation of scientific foundation models and intelligent agents. The toolkit is open-sourced and actively maintained to foster community-driven development and progress in AI4Science.

cs.AI↗