arXiv subjects

Yanlin Li

Publications and source records attributed to Yanlin Li.

16 recordsLinked to original sources

arXiv2026

VitronEdit: Universal Visual Editing across Images, Videos, and 3D Scenes

Recent advances in unified visual editing have begun to consolidate image and video editing within a single model, while 3D scene editing remains largely separate and constrained by scarce paired training data. Given the rich image and video editing supervision now available and the common editing semantics shared across modalities, a natural question is whether this knowledge can be transferred to data-scarce 3D editing. We present VitronEdit, a universal framework for image, video, and 3D scene editing. VitronEdit adopts a shared sequence representation and progressive multimodal training to transfer editing knowledge across modalities. We further introduce geometry-aware post-training, which leverages a pretrained geometry reconstruction model to provide depth and camera-pose supervision. Surprisingly, applying geometric supervision only to 3D samples yields the best overall performance, while image and video editing also improve despite receiving no direct geometric supervision, suggesting that geometric regularization transfers through the shared backbone. Our analysis further shows that image and video training mainly transfers editing semantics to 3D, whereas 3D supervision primarily improves multi-view consistency. The project page: https://vitron-llm.github.io/VitronEdit/.

cs.CV↗

arXiv2026

A Two-Stage Optimization Framework for Real-World Large-Scale University Course Timetabling Based on Orthogonalized Hierarchical Modeling and Conflict-Driven ALNS Repair

University course timetabling is a large-scale combinatorial optimization problem involving strongly coupled temporal and resource constraints. This paper proposes a two-stage optimization framework for real-world large-scale university course timetabling, integrating orthogonalized hierarchical modeling with conflict-driven adaptive large neighborhood search (ALNS) repair. The first stage performs global course-level optimization, while the second identifies residual conflicts through a conflict graph and localizes them into connected conflict regions. The Course--Occurrence--Session representation provides the structural basis for multi-granularity repair, which employs Cluster-, Course-, Occurrence-, and Session-level destroy operators within the ALNS framework, followed by fine-grained greedy repair. The framework is evaluated on a real-world university timetabling instance involving 3,073 course groups, 8,015 scheduling events, 998 student classes, and 1,511 instructors. Ablation experiments demonstrate the effectiveness of hierarchical repair granularity and adaptive operator selection. Across ten independent runs, the ALNS + Greedy framework achieves a mean residual conflict count of 18.50, compared with 172 for deterministic Greedy-only repair under the same initial scheduling conditions, representing an 89.24\% reduction. These results support the effectiveness of the proposed conflict-driven hierarchical repair framework for large-scale university course timetabling.

math.OC↗

arXiv2026

Rethinking Scientific Discovery in the Agentic Era

Artificial intelligence has advanced scientific discovery, but most AI4Science systems remain fragmented tools that rely on humans to coordinate problem formulation, literature grounding, model use, simulation, validation, and knowledge reuse. This paper presents \textbf{SCION (Scientific Collaborative Innovation with Agentic Organizational Nexus)}, an agentic scientific operating system that acts as an \textbf{organizational nexus}. Through a Science Agent serving as a \textbf{Meta-Harness}, SCION connects scientific tasks, tools, agents, artifacts, and memory, transforming research into an executable, auditable, and reusable operational process. At its core is the \textbf{Research Execution Plan (REP)}, which compiles high-level scientific intent into staged objectives, dependencies, verification checkpoints, tool requirements, expected artifacts, and fallback conditions. SCION further integrates hierarchical multi-agent execution, profile-driven specialization, selective context construction, governed delegation, and layered epistemic memory to support long-horizon scientific work. We formulate discovery under SCION as \textbf{Target-conditioned Inverse Search} and extend it to hidden-target settings through batch active search under finite experimental budgets. Applications in materials analysis, molecule design, and protein or antibody screening, together with experiments on scientific reading, idea generation, molecule generation, and antibody screening, show that SCION outperforms existing autonomous research-agent baselines, especially in decomposition, verification, refinement, and memory reuse. Overall, SCION shifts AI from isolated tools toward a coordinated operational layer for traceable and reusable scientific innovation.

cs.CL↗

arXiv2026

Toward Recursive Self-Improvement in Physical AI

Recursive self-improvement (RSI) describes a system that improves not only its behavior but also the process by which it improves. Recent digital agents test this idea by rewriting their own code and checking each version with executable tests. Bringing these mechanisms into the physical world requires learning from partial observations through costly trials, while accounting for changing operating conditions and potentially irreversible consequences. We position RSI within robot learning and trace pathways from improving robot capabilities to improving the processes that generate them. We examine how physical experience can strengthen these improvement mechanisms and when the resulting gains can support further self-improvement of the mechanisms themselves. Gains carry forward only when feedback addresses remaining learning needs, the receiving component can use the change, its value survives changes in tasks or hardware, and continued practice remains safe and affordable. Existing studies, concentrated in manipulation, support behavioral improvement, benefits to later learning, and procedure improvement. Some systems also change the improver's knowledge or tools, but the physical studies reviewed here have yet to isolate gains attributable to recursion. We synthesize these findings into evaluation principles and research directions for machines that, trial by trial, safely improve the way they improve.

cs.RO↗

arXiv2026

Omni-IO Skills: Harnessing Your Agent Omni-Native

General-purpose agents can plan, reason, and act over long horizons, yet their production capabilities remain fragmented across text, images, audio, video, documents, 3D assets, and code. Extending a foundation model to additional modalities ties capability growth to costly model updates, while assembling specialist models and tools leaves unresolved how procedures, dependencies, intermediate assets, and cross-turn revisions should be coordinated. We present Omni-IO Skills, a plug-and-play Agent Harness that makes existing agents omni-native through hierarchical Skills, a standardized multimodal execution interface, dependency-aware orchestration, and a persistent Asset Registry. Multi-asset workflows are represented as Declare Execution Graphs, which schedule independent operations concurrently and register successful outputs for downstream and cross-turn reuse across replaceable execution backends. Its 27 Skills cover 38 representative tasks spanning seven artifact modalities and four capability families: understanding, generation, reasoning, and retrieval. On UniM-90, the harness raises the input-support rates of GPT-5.6 Sol and Claude Sonnet 5 from 40.00% and 38.89% to 100%, while increasing relative Semantic--Quality Coupled Score from 26.99 to 74.94 and from 27.82 to 77.78, respectively; Strict Structure Score reaches 100.00 and 99.78. These results establish harness-level capability composition as a practical route to broad, evolvable Omni systems without changing the host agent's reasoning core.

cs.CL↗

arXiv2026

AI for Games in the Foundation Model Era

Foundation models, alongside advances in learned game-world models, are reshaping AI across the game lifecycle. Beyond playing games, recent systems model players and game dynamics, support design and development, adapt player-facing experiences at runtime, and evaluate resulting artifacts. Yet these directions have evolved largely separately, obscuring which capabilities transfer across settings and which remain tied to particular games, engines, interfaces, or player populations. We organize the literature into six roles according to the immediate use of AI output: playing and acting; modeling players and games; designing games; building and maintaining games; generating and adapting at runtime; and testing and evaluating games. For each role, we examine what structure is supplied by the game or workflow, what AI learns or produces, which capabilities and artifacts transfer across settings and roles, and what evidence supports the claims. We identify cross-role connections: trajectories train world models, learned environments provide experience for agents, design specifications drive executable implementations, and play or testing feedback guides revision. However, control schemes, rules, engine interfaces, state representations, and player contexts often remain setting-specific, so downstream claims require validation in the target setting. Evaluation is most standardized for bounded game playing and selected learned environments, while persistent state in learned worlds, repeated software revision, validated player modeling, sustained runtime adaptation, and representative automated testing remain less established. The central challenge is to reuse or transfer outputs and capabilities across roles while re-establishing evidence for effectiveness in the game-specific contexts where they are used.

cs.AI↗

arXiv2026

UniM: A Unified Any-to-Any Interleaved Multimodal Benchmark

In real-world multimodal applications, systems usually need to comprehend arbitrarily combined and interleaved multimodal inputs from users, while also generating outputs in any interleaved multimedia form. This capability defines the goal of any-to-any interleaved multimodal learning under a unified paradigm of understanding and generation, posing new challenges and opportunities for advancing Multimodal Large Language Models (MLLMs). To foster and benchmark this capability, this paper introduces the UniM benchmark, the first Unified Any-to-Any Interleaved Multimodal dataset. UniM contains 31K high-quality instances across 30 domains and 7 representative modalities: text, image, audio, video, document, code, and 3D, each requiring multiple intertwined reasoning and generation capabilities. We further introduce the UniM Evaluation Suite, which assesses models along three dimensions: Semantic Correctness & Generation Quality, Response Structure Integrity, and Interleaved Coherence. In addition, we propose UniMA, an agentic baseline model equipped with traceable reasoning for structured interleaved generation. Comprehensive experiments demonstrate the difficulty of UniM and highlight key challenges and directions for advancing unified any-to-any multimodal intelligence. The project page is https://any2any-mllm.github.io/unim.

cs.CV↗

arXiv2026

MIST: Towards Multi-dimensional Implicit BiaS Evaluation of LLMs for Theory of Mind

Theory of Mind (ToM) in Large Language Models (LLMs) refers to the model's ability to infer the mental states of others, with failures in this ability often manifesting as systemic implicit biases. Assessing this challenge is difficult, as traditional direct inquiry methods are often met with refusal to answer and fail to capture its subtle and multidimensional nature. Therefore, we propose MIST, which reconceptualizes the content model of stereotypes into multidimensional failures of ToM, specifically in the domains of competence, sociability, and morality. The framework introduces two indirect tasks. The Word Association Bias Test (WABT) assesses implicit lexical associations, while the Affective Attribution Test (AAT) measures implicit emotional tendencies, aiming to uncover latent stereotypes without triggering model avoidance. Through extensive experimentation on eight state-of-the-art LLMs, our framework demonstrates the ability to reveal complex bias structures and improved robustness. All data and code will be released.

cs.CL↗

arXiv2025

Distributed Rotary Coverage Control of Multi-Agent Systems in Uncertain Environments

It is always a challenging task for multi-agent systems to achieve efficient and robust coverage in uncertain environments. The absence of global positioning information on the uncertain environment introduces significant complexity to the spatially distributed design of coverage control algorithms. To address this issue, this paper proposes a coverage control formulation based on beacon-free rotary pointer partition mechanism. A partition dynamics is designed to enable the asymptotical consensus of multi-agent reference points, as well as the workload-balanced subdivision of coverage region. On this basis, a distributed coverage control algorithm is developed to drive each agent toward the optimal deployment of their respective subregions, thereby minimizing the coverage cost. Simulation results demonstrate that the proposed coverage control method can significantly improve overall coverage efficiency with workload balance among agents, and exhibit strong adaptability and robustness in uncertain environments.

math.OC↗

arXiv2025

Evaluating MLLMs with Multimodal Multi-image Reasoning Benchmark

With enhanced capabilities and widespread applications, Multimodal Large Language Models (MLLMs) are increasingly required to process and reason over multiple images simultaneously. However, existing MLLM benchmarks focus either on single-image visual reasoning or on multi-image understanding tasks with only final-answer evaluation, leaving the reasoning capabilities of MLLMs over multi-image inputs largely underexplored. To address this gap, we introduce the $\textbf{Multimodal Multi-image Reasoning Benchmark (MMRB)}$, the first benchmark designed to evaluate structured visual reasoning across multiple images. MMRB comprises $\textbf{92 sub-tasks}$ covering spatial, temporal, and semantic reasoning, with multi-solution, CoT-style annotations generated by GPT-4o and refined by human experts. A derivative subset is designed to evaluate multimodal reward models in multi-image scenarios. To support fast and scalable evaluation, we propose a sentence-level matching framework using open-source LLMs. Extensive baseline experiments on $\textbf{40 MLLMs}$, including 9 reasoning-specific models and 8 reward models, demonstrate that open-source MLLMs still lag significantly behind commercial MLLMs in multi-image reasoning tasks. Furthermore, current multimodal reward models are nearly incapable of handling multi-image reward ranking tasks.

cs.CV↗

arXiv2025

Gradient estimation of a generalized non-linear heat type equation along Super-Perelman Ricci flow on weighted Riemannian manifolds

In this article we derive gradient estimation for positive solution of the equation \begin{equation*} (\partial_t-Δ_f)u = A(u)p(x,t) + B(u)q(x,t) + \mathcal{G}(u) \end{equation*} on a weighted Riemannian manifold evolving along the $(k,m)$ super Perelman-Ricci flow \begin{equation*} \frac{\partial g}{\partial t}(x,t)+2Ric_f^m(g)(x,t)\ge -2kg(x,t). \end{equation*} As an application of gradient estimation we derive a Harnack type inequality along with a Liouville type theorem.

math.DG↗

arXiv2024

Boundary Value Problem and Discrete Schwarz-Pick Lemma for Generalized Hyperbolic Circle Packings

In 1991, Beardon and Stephenson [2] generalized the classical Schwarz-Pick lemma in hyperbolic geometry to the discrete Schwarz-Pick lemma for Andreev circle packings. This paper continues to investigate the discrete Schwarz-Pick lemma for generalized circle packings (including circle, horocycle or hypercycle) in hyperbolic background geometry. Since the discrete Schwarz-Pick lemma is to compare some geometric quantities of two generalized circle packings with different boundary values, we first show the existence and rigidity of generalized circle packings with boundary values, and then we introduce the method of combinatorial Calabi flows to find the generalized circle packings with boundary values. Moreover, motivated by the method of He [21], we propose the maximum principle for generalized circle packings. Finally, we use the maximum principle to prove the discrete Schwarz-Pick lemma for generalized circle packings.

math.DG↗

arXiv2022

Timeliness of Information for Computation-intensive Status Updates in Task-oriented Communications

Moving beyond just interconnected devices, the increasing interplay between communication and computation has fed the vision of real-time networked control systems. To obtain timely situational awareness, IoT devices continuously sample computation-intensive status updates, generate perception tasks and offload them to edge servers for processing. In this sense, the timeliness of information is considered as one major contextual attribute of status updates. In this paper, we derive the closed-form expressions of timeliness of information for computation offloading at both edge tier and fog tier, where two stage tandem queues are exploited to abstract the transmission and computation process. Moreover, we exploit the statistical structure of Gauss-Markov process, which is widely adopted to model temporal dynamics of system states, and derive the closed-form expression for process-related timeliness of information. The obtained analytical formulas explicitly characterize the dependency among task generation, transmission and execution, which can serve as objective functions for system optimization. Based on the theoretical results, we formulate a computation offloading optimization problem at edge tier, where the timeliness of status updates is minimized among multiple devices by joint optimization of task generation, bandwidth allocation, and computation resource allocation. An iterative solution procedure is proposed to solve the formulated problem. Numerical results reveal the intertwined relationship among transmission and computation stages, and verify the necessity of factoring in the task generation process for computation offloading strategy design.

cs.DC↗

arXiv2021

Rooted quasi-Stirling permutations of general multisets

Given a general multiset $\mathcal{M}=\{1^{m_1},2^{m_2},\ldots,n^{m_n}\}$, where $i$ appears $m_i$ times, a multipermutation $π$ of $\mathcal{M}$ is called {\em quasi-Stirling}, if it contains no subword of the form $abab$ with $a\neq b$. We designate exactly one entry of $π$, say $k\in \mathcal{M}$, which is not the leftmost entry among all entries with the same value, by underlining it in $π$, and we refer to the pair $(π,k)$ as a quasi-Stirling multipermutation of $\mathcal{M}$ rooted at $k$. By introducing certain vertex and edge labeled trees, we give a new bijective proof of an identity due to Yan, Yang, Huang and Zhu, which links the enumerator of rooted quasi-Stirling multipermutations by the numbers of ascents, descents, and plateaus, with the exponential generating function of the {\em bivariate Eulerian polynomials}. This identity can be viewed as a natural extension of Elizalde's result on $k$-quasi-Stirling permutations, and our bijective approach to proving it enables us to: (1) prove bijectively a Carlitz type identity involving quasi-Stirling polynomials on multisets that was first obtained by Yan and Zhu; (2) confirm a recent partial $γ$-positivity conjecture due to Lin, Ma and Zhang, and find a combinatorial interpretation of the $γ$-coefficients in terms of two new statistics defined on quasi-Stirling multipermutations called sibling descents and double sibling descents.

math.CO↗

arXiv2020

EEMC: Embedding Enhanced Multi-tag Classification

The recently occurred representation learning make an attractive performance in NLP and complex network, it is becoming a fundamental technology in machine learning and data mining. How to use representation learning to improve the performance of classifiers is a very significance research direction. We using representation learning technology to map raw data(node of graph) to a low-dimensional feature space. In this space, each raw data obtained a lower dimensional vector representation, we do some simple linear operations for those vectors to produce some virtual data, using those vectors and virtual data to training multi-tag classifier. After that we measured the performance of classifier by F1 score(Macro% F1 and Micro% F1). Our method make Macro F1 rise from 28 % - 450% and make average F1 score rise from 12 % - 224%. By contrast, we trained the classifier directly with the lower dimensional vector, and measured the performance of classifiers. We validate our algorithm on three public data sets, we found that the virtual data helped the classifier greatly improve the F1 score. Therefore, our algorithm is a effective way to improve the performance of classifier. These result suggest that the virtual data generated by simple linear operation, in representation space, still retains the information of the raw data. It's also have great significance to the learning of small sample data sets.

cs.LG↗

arXiv2015

Joint Channel Assignment and Opportunistic Routing for Maximizing Throughput in Cognitive Radio Networks

In this paper, we consider the joint opportunistic routing and channel assignment problem in multi-channel multi-radio (MCMR) cognitive radio networks (CRNs) for improving aggregate throughput of the secondary users. We first present the nonlinear programming optimization model for this joint problem, taking into account the feature of CRNs-channel uncertainty. Then considering the queue state of a node, we propose a new scheme to select proper forwarding candidates for opportunistic routing. Furthermore, a new algorithm for calculating the forwarding probability of any packet at a node is proposed, which is used to calculate how many packets a forwarder should send, so that the duplicate transmission can be reduced compared with MAC-independent opportunistic routing & encoding (MORE) [11]. Our numerical results show that the proposed scheme performs significantly better that traditional routing and opportunistic routing in which channel assignment strategy is employed.

cs.NI↗