arXiv subjects

Bo Li

Publications and source records attributed to Bo Li.

At least 217 records · Page 12Linked to original sources

arXiv2026

UniG2U-Bench: Do Unified Models Advance Multimodal Understanding?

Unified multimodal models have recently demonstrated strong generative capabilities, yet whether and when generation improves understanding remains unclear. Existing benchmarks lack a systematic exploration of the specific tasks where generation facilitates understanding. To this end, we introduce UniG2U-Bench, a comprehensive benchmark categorizing generation-to-understanding (G2U) evaluation into 7 regimes and 30 subtasks, requiring varying degrees of implicit or explicit visual transformations. Extensive evaluation of over 30 models reveals three core findings: 1) Unified models generally underperform their base Vision-Language Models (VLMs), and Generate-then-Answer (GtA) inference typically degrades performance relative to direct inference. 2) Consistent enhancements emerge in spatial intelligence, visual illusions, or multi-round reasoning subtasks, where enhanced spatial and shape perception, as well as multi-step intermediate image states, prove beneficial. 3) Tasks with similar reasoning structures and models sharing architectures exhibit correlated behaviors, suggesting that generation-understanding coupling induces class-consistent inductive biases over tasks, pretraining data, and model architectures. These findings highlight the necessity for more diverse training data and novel paradigms to fully unlock the potential of unified multimodal modeling.

cs.CV↗

arXiv2026

Incomplete Multi-Label Image Recognition by Co-learning Semantic-Aware Features and Label Recovery

Multi-label image recognition with incomplete labels is a challenging yet vital task in computer vision, which faces two fundamental challenges: learning semantic-aware features and recovering missing labels. In this paper, we propose a Co-learning framework for Semantic-aware features and Label recovery (CSL), designed to address both challenges in a unified learning paradigm. Specifically, we develop a semantic-related feature learning module that captures robust semantic-related representations by discovering semantic information and label correlations. Furthermore, a semantic-guided feature enhancement module is introduced to generate highly discriminative semantic-aware features by effectively aligning visual and semantic spaces. Finally, we present a collaborative learning framework that integrates semantic-aware feature learning with label recovery. This framework not only dynamically enhances the discriminability of semantic-aware features but also adaptively infers and recovers missing labels, thereby forming a mutually reinforcing mechanism between the two processes. Extensive experiments on three widely used public datasets (MS-COCO, VOC2007, and NUS-WIDE) demonstrate that CSL outperforms state-of-the-art methods for incomplete multi-label image recognition.

cs.CV↗

arXiv2026

The derivative of the fractional discrete Laplacian is an exotic Riesz potential

Let $Δ_{N}$ be the multidimensional discrete Laplacian on $\mathbb{Z}^N$ ($N\ge1$). In this note, we prove that, when $N=1$, the right hand derivative of $(-Δ_1)^s$ at $0$ is an exotic discrete Riesz potential (namely, the endpoint case: the order is 0) in Stein-Wainger sense (J. Anal. Math. 2000), and when $N\ge 2$, the corresponding derivative is also an exotic discrete Riesz potential with an additional corrector. A similar conclusion for the left hand derivative case is also considered. All results obtained in this note extend the logarithmic Laplacian of Chen-Weth (Comm. PDEs. 2019) to the discrete setting.

math.AP↗

arXiv2026

Time-Aware One Step Diffusion Network for Real-World Image Super-Resolution

Diffusion-based real-world image super-resolution (Real-ISR) methods have demonstrated impressive performance.To achieve efficient Real-ISR, many works employ Variational Score Distillation (VSD) to distill pre-trained stable-diffusion (SD) model for one-step SR with a fixed timestep. However, since SD will perform different generative priors at different timesteps, a fixed timestep is difficult for these methods to fully leverage the generative priors in SD, leading to suboptimal performance.To address this, we propose a \textbf{T}ime-\textbf{A}ware one-step \textbf{D}iffusion Network for Real-ISR (\textbf{TADSR}). We first introduce a Time-Aware VAE Encoder, which projects the same image into different latent features based on timesteps.Through joint dynamic variation of timesteps and latent features, the student model can better align with the input pattern distribution of the pre-trained SD, thereby enabling more effective utilization of SD's generative capabilities.To better activate the generative prior of SD at different timesteps, we propose a Time-Aware VSD loss that bridges the timesteps of the student model and those of the teacher model, thereby producing more consistent generative prior guidance conditioned on timesteps. Additionally, though utilizing the generative prior in SD at different timesteps, our method can naturally achieve \textbf{controllable trade-offs between fidelity and realism} by changing the timestep.Experimental results demonstrate that our method achieves both state-of-the-art performance and controllable SR results with only a single step. The source codes are released at https://github.com/zty557/TADSR

eess.IV↗

arXiv2026

Lower Bias, Higher Welfare: How Creator Competition Reshapes Bias-Variance Tradeoff in Recommendation Platforms?

Understanding the bias-variance tradeoff in user representation learning is essential for improving recommendation quality in modern content platforms. While well studied in static settings, this tradeoff becomes significantly more complex when content creators strategically adapt to platform incentives. To analyze how such competition reshapes the tradeoff for maximizing user welfare, we introduce the Content Creator Competition with Bias-Variance Tradeoff framework, a tractable game-theoretic model that captures the platform's decision on regularization strength in user feature estimation. We derive and compare the platform's optimal policy under two key settings: a non-strategic baseline with fixed content and a strategic environment where creators compete in response to the platform's algorithmic design. Our theoretical analysis in a stylized model shows that, compared to the non-strategic environment, content creator competition shifts the platform's optimal policy toward weaker regularization, thereby favoring lower bias in the bias-variance tradeoff. To validate and assess the robustness of these insights beyond the stylized setting, we conduct extensive experiments on both synthetic and real-world benchmark datasets. The empirical results consistently support our theoretical conclusion: in strategic environments, reducing bias leads to higher user welfare. These findings offer practical implications for the design of real-world recommendation algorithms in the presence of content creator competition.

cs.GT↗

arXiv2026

Micro-expression Recognition Based on Dual-branch Feature Extraction and Fusion

Micro-expressions, characterized by transience and subtlety, pose challenges to existing optical flow-based recognition methods. To address this, this paper proposes a dual-branch micro-expression feature extraction network integrated with parallel attention. Key contributions include: 1) a residual network designed to alleviate gradient anishing and network degradation; 2) an Inception network constructed to enhance model representation and suppress interference from irrelevant regions; 3) an adaptive feature fusion module developed to integrate dual-branch features. Experiments on the CASME II dataset demonstrate that the proposed method achieves 74.67% accuracy, outperforming LBP-TOP (by 11.26%), MSMMT (by 3.36%), and other comparative methods.

cs.CV↗

arXiv2026

OneVision-Encoder: Codec-Aligned Sparsity as a Foundational Principle for Multimodal Intelligence

Hypothesis. Artificial general intelligence is, at its core, a compression problem. Effective compression demands resonance: deep learning scales best when its architecture aligns with the fundamental structure of the data. These are the fundamental principles. Yet, modern vision architectures have strayed from these truths: visual signals are highly redundant, while discriminative information, the surprise, is sparse. Current models process dense pixel grids uniformly, wasting vast compute on static background rather than focusing on the predictive residuals that define motion and meaning. We argue that to solve visual understanding, we must align our architectures with the information-theoretic principles of video, i.e., Codecs. Method. OneVision-Encoder encodes video by compressing predictive visual structure into semantic meaning. By adopting Codec Patchification, OV-Encoder abandons uniform computation to focus exclusively on the 3.1%-25% of regions rich in signal entropy. To unify spatial and temporal reasoning under irregular token layouts, OneVision-Encoder employs a shared 3D RoPE and is trained with a large-scale cluster discrimination objective over more than one million semantic concepts, jointly capturing object permanence and motion dynamics. Evidence. The results validate our core hypothesis: efficiency and accuracy are not a trade-off; they are positively correlated. When integrated into LLM, it consistently outperforms strong vision backbones such as Qwen3-ViT and SigLIP2 across 16 image, video, and document understanding benchmarks, despite using substantially fewer visual tokens and pretraining data. Notably, on video understanding tasks, OV-Encoder achieves an average improvement of 4.1% over Qwen3-ViT. Codec-aligned, patch-level sparsity is a foundational principle, enabling OV-Encoder as a scalable engine for next-generation visual generalists.

cs.CV↗

arXiv2026

Tokenization, Fusion and Decoupling: Bridging the Granularity Mismatch Between Large Language Models and Knowledge Graphs

Leveraging Large Language Models (LLMs) for Knowledge Graph Completion (KGC) is promising but hindered by a fundamental granularity mismatch. LLMs operate on fragmented token sequences, whereas entities are the fundamental units in knowledge graphs (KGs) scenarios. Existing approaches typically constrain predictions to limited candidate sets or align entities with the LLM's vocabulary by pooling multiple tokens or decomposing entities into fixed-length token sequences, which fail to capture both the semantic meaning of the text and the structural integrity of the graph. To address this, we propose KGT, a novel framework that uses dedicated entity tokens to enable efficient, full-space prediction. Specifically, we first introduce specialized tokenization to construct feature representations at the level of dedicated entity tokens. We then fuse pre-trained structural and textual features into these unified embeddings via a relation-guided gating mechanism, avoiding training from scratch. Finally, we implement decoupled prediction by leveraging independent heads to separate and combine semantic and structural reasoning. Experimental results show that KGT consistently outperforms state-of-the-art methods across multiple benchmarks.

cs.CL↗

arXiv2026

Identifiability of Treatment Effects with Unobserved Spatially Varying Confounders

The study of causal effects in the presence of unmeasured spatially varying confounders has garnered increasing attention. However, a general framework for identifiability, which is critical for reliable causal inference from observational data, has yet to be advanced. In this paper, we study a linear model with various parametric model assumptions on the covariance structure between the unmeasured confounder and the exposure of interest. We establish identifiability of the treatment effect for many commonly 20 used spatial models for both discrete and continuous data, under mild conditions on the structure of observation locations and the exposure-confounder association. We also emphasize models or scenarios where identifiability may not hold, under which statistical inference should be conducted with caution.

stat.ME↗

arXiv2026

BrowseComp-$V^3$: A Visual, Vertical, and Verifiable Benchmark for Multimodal Browsing Agents

Multimodal large language models (MLLMs), equipped with increasingly advanced planning and tool-use capabilities, are evolving into autonomous agents capable of performing multimodal web browsing and deep search in open-world environments. However, existing benchmarks for multimodal browsing remain limited in task complexity, evidence accessibility, and evaluation granularity, hindering comprehensive and reproducible assessments of deep search capabilities. To address these limitations, we introduce BrowseComp-$V^3$, a novel benchmark consisting of 300 carefully curated and challenging questions spanning diverse domains. The benchmark emphasizes deep, multi-level, and cross-modal multi-hop reasoning, where critical evidence is interleaved across textual and visual modalities within and across web pages. All supporting evidence is strictly required to be publicly searchable, ensuring fairness and reproducibility. Beyond final-answer accuracy, we incorporate an expert-validated, subgoal-driven process evaluation mechanism that enables fine-grained analysis of intermediate reasoning behaviors and systematic characterization of capability boundaries. In addition, we propose OmniSeeker, a unified multimodal browsing agent framework integrating diverse web search and visual perception tools. Comprehensive experiments demonstrate that even state-of-the-art models achieve only 36% accuracy on our benchmark, revealing critical bottlenecks in multimodal information integration and fine-grained perception. Our results highlight a fundamental gap between current model capabilities and robust multimodal deep search in real-world settings.

cs.AI↗

arXiv2026

AnimeAgent: Is the Multi-Agent via Image-to-Video models a Good Disney Storytelling Artist?

Custom Storyboard Generation (CSG) aims to produce high-quality, multi-character consistent storytelling. Current approaches based on static diffusion models, whether used in a one-shot manner or within multi-agent frameworks, face three key limitations: (1) Static models lack dynamic expressiveness and often resort to "copy-paste" pattern. (2) One-shot inference cannot iteratively correct missing attributes or poor prompt adherence. (3) Multi-agents rely on non-robust evaluators, ill-suited for assessing stylized, non-realistic animation. To address these, we propose AnimeAgent, the first Image-to-Video (I2V)-based multi-agent framework for CSG. Inspired by Disney's "Combination of Straight Ahead and Pose to Pose" workflow, AnimeAgent leverages I2V's implicit motion prior to enhance consistency and expressiveness, while a mixed subjective-objective reviewer enables reliable iterative refinement. We also collect a human-annotated CSG benchmark with ground-truth. Experiments show AnimeAgent achieves SOTA performance in consistency, prompt fidelity, and stylization.

cs.CV↗

arXiv2026

FAMOSE: A ReAct Approach to Automated Feature Discovery

Feature engineering remains a critical yet challenging bottleneck in machine learning, particularly for tabular data, as identifying optimal features from an exponentially large feature space traditionally demands substantial domain expertise. To address this challenge, we introduce FAMOSE (Feature AugMentation and Optimal Selection agEnt), a novel framework that leverages the ReAct paradigm to autonomously explore, generate, and refine features while integrating feature selection and evaluation tools within an agent architecture. To our knowledge, FAMOSE represents the first application of an agentic ReAct framework to automated feature engineering, especially for both regression and classification tasks. Extensive experiments demonstrate that FAMOSE is at or near the state-of-the-art on classification tasks (especially tasks with more than 10K instances, where ROC-AUC increases 0.23% on average), and achieves the state-of-the-art for regression tasks by reducing RMSE by 2.0% on average, while remaining more robust to errors than other algorithms. We hypothesize that FAMOSE's strong performance is because ReAct allows the LLM context window to record (via iterative feature discovery and evaluation steps) what features did or did not work. This is similar to a few-shot prompt and guides the LLM to invent better, more innovative features. Our work offers evidence that AI agents are remarkably effective in solving problems that require highly inventive solutions, such as feature engineering.

cs.LG↗

arXiv2026

Large-Time Asymptotics for Heat and Fractional Heat Equations on the Lattice and General Finite Subgraphs

In this paper, we study large-time asymptotics for heat and fractional heat equations in two discrete settings: the full lattice $\mathbb Z^d$ and finite connected subgraphs with Dirichlet boundary condition. These results provide a unified discrete theory of long-time asymptotics for local and nonlocal diffusions. For $d\ge1$ and $s\in(0,1]$, we consider on $\mathbb Z^d$ the Cauchy problem \[ \partial_t u+(-Δ)^s u=0,\qquad u(0)=u_0\in \ell^1(\mathbb Z^d), \] and derive a precise first-order asymptotic expansion toward the lattice fractional heat kernel $G_t^{(s)}$. The main technical input is a pair of sharp translation-increment bounds for $G_t^{(s)}$: a pointwise estimate and an $\ell^1$-estimate. As consequences, under finite first moment we obtain the optimal decay rate $t^{-1/(2s)}$ in $\ell^p$-asymptotics ($1\le p\le\infty$), and we prove sharpness by explicit shifted-kernel examples. Without moment assumptions, we still establish convergence in the full $\ell^1$-class, and we show that no universal quantitative rate can hold in general. We also analyze fractional Dirichlet diffusion on finite connected subgraphs (restricted fractional setting, including $s=1$ as the local case). In this finite-dimensional framework, solutions admit spectral decomposition and exhibit exponential large-time behavior governed by the principal eigenvalue and the spectral gap. In addition, we study positivity improving properties of the associated semigroups for both the lattice and Dirichlet evolutions.

math.AP↗

arXiv2026

Three-dimensional kink modes in solar coronal slabs: group velocities and their implications for impulsively excited waves

Little attention has been paid to group velocities of three-dimensional (3D) MHD waves in solar coronal seismology. This study aims to present a rather comprehensive examination on the group velocities of trapped 3D kink modes in coronal slabs, emphasizing the connection of mode analysis to both mode characterization and impulsively excited 3D kink waves. We work in linear, ideal, pressureless MHD, and take the equilibrium slab to be symmetrically structured only in one transverse direction. The dispersion relation is numerically solved, with the results understood by making in-depth analytical progress. We address both the transverse fundamental and its first overtone. We develop a three-subgroup scheme for categorizing 3D kink modes on the plane spanned by the axial and out-of-plane wavenumbers. The group ($\vec{v}_{\rm gr}$) and phase velocities ($\vec{v}_{\rm ph}$) sit on the same side of the equilibrium magnetic field ($\vec{B}_0$) for the ``$\vec{B}_0$-same-side A'' and ``$\vec{B}_0$-same-side F'' subgroups, which are further discriminated by the directional similarity of $\vec{v}_{\rm gr}$ and $\vec{B}_0$. The ``$\vec{B}_0$-straddling'' subgroup is peculiar in that $\vec{v}_{\rm gr}$ and $\vec{v}_{\rm ph}$ lie astride $\vec{B}_0$, a feature that cannot be found for waves in unbounded uniform media in pressureless MHD. This ``$\vec{B}_0$-straddling'' subgroup pertains to both the fundamental and its overtones. We further place our results in the context of impulsive waves, employing the method of stationary phase to predict the large-time wavefront morphology in the plane of symmetry of the equilibrium slab. Wavefronts directed toward $\vec{B}_0$ derive exclusively from ``$\vec{B}_0$-straddling'' modes, and are confined to narrow sectors.

astro-ph.SR↗

arXiv2026

Deterministic Switching of the Néel Vector by Asymmetric Spin Torque

Néel vector, the order parameter of collinear antiferromagnets, serves as a state variable in associated antiferromagnetic (AFM) spintronic devices to encode information. A deterministic switching of Néel vector is crucial for the write-in operation, which, however, remains a challenging problem in AFM spintronics. Here we demonstrate, based on analytical derivation and macro-spin simulations, that Néel vector switching can be generally achieved via a current-induced spin torque, provided the spin accumulations responsible for this torque are non-identical between opposite sublattices. This condition occurs widely in AFM films, as symmetry equivalence between sublattice-dependent spin accumulations is usually absent, allowing unequal spin accumulations induced by Edelstein effect or a spin current. Unlike previously studied spin torques induced by uniform or staggered spin accumulations -- where either the field-like or damping-like component dominates exclusively -- the asymmetric spin torque features cooperative contributions from both components, leading to Néel vector dynamics that are fundamentally distinct from previous expectations. Crucially, the static states stabilized by the asymmetric spin torque enable versatile Néel vector switching strategies -- field-free spin-transfer torque switching during current application, as well as field-free or field-assisted spin-orbit torque switching after the current pulse -- demonstrating that established spin torque techniques from ferromagnetic spintronics can be directly adapted to AFM systems, a capability absent in previous theoretical frameworks. Our work establishes a general mechanism for current-induced Néel vector switching, which is in principle feasible for all collinear antiferromagnets, and thus paves the route to realize efficient writing in AFM spintronics.

cond-mat.mtrl-sci↗

arXiv2026

The Effect of Gravitational Stratification on Kink Oscillations in Curved Coronal Loops

Kink oscillation frequency is a key parameter for coronal seismology. It is still unclear how gravitational stratification affects the kink frequency in curved coronal loops. This work aims to investigate the effect of gravitational stratification on the frequency of kink oscillations in curved coronal loops and discuss their seismological potential. We conduct numerical computations within the ideal MHD framework to study different kink polarizations and harmonics in a curved, gravitationally stratified coronal loop. The oscillation frequencies derived from the Lagrangian displacement are compared with the WKB approximation. For the vertically polarized fundamental mode, the oscillation frequency deviates from the WKB approximation by about 18\% in the current numerical setup. Nevertheless, the oscillation frequency closely matches the local Alfvén frequency near the loop apex. On the other hand, the frequency of the horizontally polarized fundamental mode exhibits only a 7\% deviation in our current model from the WKB approximation and closely matches the local Alfvén frequency near one quarter of the loop. For the first overtones, the frequencies for both polarizations can be well described by the WKB approximation. The frequency of vertically polarized fundamental kink modes can be predicted by the local Alfvén frequency near the loop apex. In contrast, the WKB approximation remains highly reliable for estimating the frequency of horizontally polarized fundamental modes and first overtones, which is also well described by the local Alfvén frequency near one quarter of the loop. These results therefore pave the way for spatially dependent coronal seismology, enabling, e.g., the probing of magnetic field strength at different locations along a coronal loop.

astro-ph.SR↗

arXiv2026

Scalable Changepoint Detection for Large Spatiotemporal Data on the Sphere

We propose a novel Bayesian framework for changepoint detection in large-scale spherical spatiotemporal data, with broad applicability in environmental and climate sciences. Our approach models changepoints as spatially dependent categorical variables using a multinomial probit model (MPM) with a latent Gaussian process, effectively capturing complex spatial correlation structures on the sphere. To handle the high dimensionality inherent in global datasets, we leverage stochastic partial differential equations (SPDE) and spherical harmonic transformations for efficient representation and scalable inference, drastically reducing computational burden while maintaining high accuracy. Through extensive simulation studies, we demonstrate the efficiency and robustness of the proposed method for changepoint estimation, as well as the significant computational gains achieved through the combined use of the MPM and truncated spectral representations of latent processes. Finally, we apply our method to global aerosol optical depth data, successfully identifying changepoints associated with a major atmospheric event.

stat.ME↗

arXiv2026

Poly-Guard: Massive Multi-Domain Safety Policy-Grounded Guardrail Dataset

As LLMs become widespread across diverse applications, concerns about the security and safety of LLM interactions have intensified. Numerous guardrail models and benchmarks have been developed to ensure LLM content safety. However, existing guardrail benchmarks are often built upon ad hoc risk taxonomies that lack a principled grounding in standardized safety policies, limiting their alignment with real-world operational requirements. Moreover, they tend to overlook domain-specific risks, while the same risk category can carry different implications across different domains. To bridge these gaps, we introduce Poly-Guard, the first massive multi-domain safety policy-grounded guardrail dataset. Poly-Guard offers: (1) broad domain coverage across eight safety-critical domains, such as finance, law, and codeGen; (2) policy-grounded risk construction based on authentic, domain-specific safety guidelines; (3) diverse interaction formats, encompassing declarative statements, questions, instructions, and multi-turn conversations; (4) advanced benign data curation via detoxification prompting to challenge over-refusal behaviors; and (5) \textbf{attack-enhanced instances} that simulate adversarial inputs designed to bypass guardrails. Based on Poly-Guard, we benchmark 19 advanced guardrail models and uncover a series of findings, such as: (1) All models achieve varied F1 scores, with many demonstrating high variance across risk categories, highlighting their limited domain coverage and insufficient handling of domain-specific safety concerns; (2) As models evolve, their coverage of safety risks broadens, but performance on common risk categories may decrease; (3) All models remain vulnerable to optimized adversarial attacks. We believe that \dataset and the unique insights derived from our evaluations will advance the development of policy-aligned and resilient guardrail systems.

cs.CR↗