arXiv subjects

Jiazhen Liu

Publications and source records attributed to Jiazhen Liu.

At least 19 recordsLinked to original sources

arXiv2026

Better, Stronger, Faster, and Broader: Structured All-Mask Prediction for MLLM-Based Segmentation

MLLM-based segmentation faces a core segmentation trilemma: high segmentation performance, preserved dialogue ability, and fast inference. Embedding-prediction methods may disrupt language modeling through pixel-level objectives, whereas next-token generation is inefficient for dense masks. We propose All-Mask Prediction, decoupling autoregressive dialogue from non-autoregressive mask prediction. Its binary instantiation, STAMP (Simultaneous Textual All-Mask Prediction), emits an in-vocabulary trigger, fuses image-aligned mask tokens with corresponding patch features, and uses hybrid attention to classify all tokens as foreground or background in one pass. It thereby combines strong referring and reasoning segmentation with preserved multimodal ability and efficient inference. However, binary masks cannot retain multiple semantic or instance identities without repeated target-specific predictions. We therefore propose Structured All-Mask Prediction and develop STAMPlus. It generates a target list with explicit IDs and optional boxes, binds these IDs to a shared multi-class mask space, and jointly predicts all targets in one non-autoregressive pass. A single unified checkpoint retains STAMP's referring and reasoning capabilities while extending to open-vocabulary semantic, instance-aware, and remote-sensing small-target segmentation, where high-resolution mask-token scaling preserves finer spatial evidence. Across these settings, STAMPlus achieves state-of-the-art segmentation performance, preserves general multimodal instruction following, and reduces 12-category latency from 13.50s for repeated STAMP inference to 5.16s. Further analyses show that accurate target cues improve segmentation and learned spatial grounding benefits look-twice reasoning. Overall, STAMPlus resolves the trilemma beyond single-target prediction.

cs.CV↗

arXiv2026

LISA: Likelihood Score Alignment for Visual-condition Controllable Generation

The prevalent dual-branch paradigm, i.e., training a side network to encode visual conditions and fusing its intermediate-layer features to a frozen pretrained main network, has shown remarkable success in visual-condition controllable generation. Despite its widespread adoption, the role of the side branch and its training efficiency remain underexplored. In this paper, we first revisit this mainstream paradigm through the lens of score-based generative modeling: 1) The main network preserves visual perceptual quality by providing a prior unconditional score. 2) The side network steers conditional control by implicitly contributing a likelihood score. Guided by this perspective, we propose LIkelihood Score Alignment (LISA), an effective regularization method that explicitly aligns the intermediate feature of the side network with an approximated likelihood score. Specifically, we first hook features from a designated layer of the side network and project them into the score latent space by a lightweight decoder. Then, we construct an approximated likelihood score target and calculate the distance between the decoder's output and this target as an additional regularization loss. Finally, we jointly optimize the side network and decoder with both standard diffusion loss and our regularization loss. Experiments across various image/video tasks, architectures, and diffusion/flow models demonstrated that LISA can not only consistently accelerate the training convergence and improve final synthetic results, but also encourage the side network's features to be more disentangled for conditional modeling with negligible additional training cost and zero extra inference cost.

cs.CV↗

arXiv2026

Fundus-R1: Training a Fundus-Reading MLLM with Knowledge-Aware Reasoning on Public Data

Fundus imaging such as CFP, OCT and UWF is crucial for the early detection of retinal anomalies and diseases. Fundus image understanding, due to its knowledge-intensive nature, poses a challenging vision-language task. An emerging approach to addressing the task is to post-train a generic multimodal large language model (MLLM), either by supervised finetuning (SFT) or by reinforcement learning with verifiable rewards (RLVR), on a considerable amount of in-house samples paired with high-quality clinical reports. However, these valuable samples are not publicly accessible, which not only hinders reproducibility but also practically limits research to few players. To overcome the barrier, we make a novel attempt to train a reasoning-enhanced fundus-reading MLLM, which we term Fundus-R1, using exclusively public datasets, wherein over 94\% of the data are annotated with only image-level labels. Our technical contributions are two-fold. First, we propose a RAG-based method for composing image-specific, knowledge-aware reasoning traces. Such auto-generated traces link visual findings identified by a generic MLLM to the image labels in terms of ophthalmic knowledge. Second, we enhance RLVR with a process reward that encourages self-consistency of the generated reasoning trace in each rollout. Extensive experiments on three fundus-reading benchmarks, i.e., FunBench, Omni-Fundus and GMAI-Fundus, show that Fundus-R1 clearly outperforms multiple baselines, including its generic counterpart (Qwen2.5-VL) and a stronger edition post-trained without using the generated traces. This work paves the way for training powerful fundus-reading MLLMs with publicly available data.

cs.CV↗

arXiv2026

Learning and Optimizing the Efficacy of Spatio-Temporal Task Allocation under Temporal and Resource Constraints

Complex multi-robot missions often require heterogeneous teams to jointly optimize task allocation, scheduling, and path planning to improve team performance under strict constraints. We formalize these complexities into a new class of problems, dubbed Spatio-Temporal Efficacy-optimized Allocation for Multi-robot systems (STEAM). STEAM builds upon trait-based frameworks that model robots using their capabilities (e.g., payload and speed), but goes beyond the typical binary success-failure model by explicitly modeling the efficacy of allocations as trait-efficacy maps. These maps encode how the aggregated capabilities assigned to a task determine performance. Further, STEAM accommodates spatio-temporal constraints, including a user-specified time budget (i.e., maximum makespan). To solve STEAM problems, we contribute a novel algorithm named Efficacy-optimized Incremental Task Allocation Graph Search (E-ITAGS) that simultaneously optimizes task performance and respects time budgets by interleaving task allocation, scheduling, and path planning. Motivated by the fact that trait-efficacy maps are difficult, if not impossible, to specify, E-ITAGS efficiently learns them using a realizability-aware active learning module. Our approach is realizability-aware since it explicitly accounts for the fact that not all combinations of traits are realizable by the robots available during learning. Further, we derive experimentally-validated bounds on E-ITAGS' suboptimality with respect to efficacy. Detailed numerical simulations and experiments using an emergency response domain demonstrate that E-ITAGS generates allocations of higher efficacy compared to baselines, while respecting resource and spatio-temporal constraints. We also show that our active learning approach is sample efficient and establishes a principled tradeoff between data and computational efficiency.

cs.RO↗

arXiv2025

WeightFlow: Learning Stochastic Dynamics via Evolving Weight of Neural Network

Modeling stochastic dynamics from discrete observations is a key interdisciplinary challenge. Existing methods often fail to estimate the continuous evolution of probability densities from trajectories or face the curse of dimensionality. To address these limitations, we presents a novel paradigm: modeling dynamics directly in the weight space of a neural network by projecting the evolving probability distribution. We first theoretically establish the connection between dynamic optimal transport in measure space and an equivalent energy functional in weight space. Subsequently, we design WeightFlow, which constructs the neural network weights into a graph and learns its evolution via a graph controlled differential equation. Experiments on interdisciplinary datasets demonstrate that WeightFlow improves performance by an average of 43.02\% over state-of-the-art methods, providing an effective and scalable solution for modeling high-dimensional stochastic dynamics.

cs.CE↗

arXiv2025

Better, Stronger, Faster: Tackling the Trilemma in MLLM-based Segmentation with Simultaneous Textual Mask Prediction

Integrating segmentation into Multimodal Large Language Models (MLLMs) presents a core trilemma: simultaneously preserving dialogue ability, achieving high segmentation performance, and ensuring fast inference. Prevailing paradigms are forced into a compromise. Embedding prediction methods introduce a conflicting pixel-level objective that degrades the MLLM's general dialogue abilities. The alternative, next-token prediction, reframes segmentation as an autoregressive task, which preserves dialogue but forces a trade-off between poor segmentation performance with sparse outputs or prohibitive inference speeds with rich ones. We resolve this trilemma with all-mask prediction, a novel paradigm that decouples autoregressive dialogue generation from non-autoregressive mask prediction. We present STAMP: Simultaneous Textual All-Mask Prediction, an MLLM that embodies this paradigm. After generating a textual response, STAMP predicts an entire segmentation mask in a single forward pass by treating it as a parallel "fill-in-the-blank" task over image patches. This design maintains the MLLM's dialogue ability by avoiding conflicting objectives, enables high segmentation performance by leveraging rich, bidirectional spatial context for all mask tokens, and achieves exceptional speed. Extensive experiments show that STAMP significantly outperforms state-of-the-art methods across multiple segmentation benchmarks, providing a solution that excels in dialogue, segmentation, and speed without compromise.

cs.CV↗

arXiv2025

Population-mobility coevolution drives spatial heterogeneity of cities

The spatial heterogeneity of cities -- the uneven distribution of population and activities -- is fundamental to urban dynamics and related to critical issues such as infrastructure overload, housing affordability, and social inequality. Despite sharing similar scaling laws of population and mobility, cities exhibit vastly different spatial patterns. This paradox call for a mechanistic explanation for the emergence of spatial heterogeneity, while existing qualitative or descriptive studies fail to capture the underlying mechanisms. Here, we propose a coupled dynamical model that describe the intra-city population-mobility coevolution, explaining spatial heterogeneity as an emergent outcome of mutual feedback between the fast-changing mobility and the slow-adapting population. Our model is validated on over 388 million records from eight diverse global cities, successfully reproduces both the statistical laws and realistic spatial patterns. We find out realistic heterogeneity emerges as a distinct stable state, intermediate between disordered homogeneity and unsustainable super-hub dominance. Moreover, we theoretically and empirically show that populated areas are predominantly shaped by coevolution strength, while the increasing distance decay leads cities through a three-phase transition of homogeneity-heterogeneity-homogeneity. Besides, functional attractiveness between areas consistently enhances the ordered heterogeneous structure. Simulations of real-world planning scenarios -- including crisis-induced lockdown, planned zone expansions, and dispersal from congested centers -- indicate that integrated population-mobility policies are more cost-effective than single interventions. Our model can provides mechanistic, high-resolution insights to rigorously inform policy design.

physics.soc-ph↗

arXiv2025

JaxRobotarium: Training and Deploying Multi-Robot Policies in 10 Minutes

Multi-agent reinforcement learning (MARL) has emerged as a promising solution for learning complex and scalable coordination behaviors in multi-robot systems. However, established MARL platforms (e.g., SMAC and MPE) lack robotics relevance and hardware deployment, leaving multi-robot learning researchers to develop bespoke environments and hardware testbeds dedicated to the development and evaluation of their individual contributions. The Multi-Agent RL Benchmark and Learning Environment for the Robotarium (MARBLER) is an exciting recent step in providing a standardized robotics-relevant platform for MARL, by bridging the Robotarium testbed with existing MARL software infrastructure. However, MARBLER lacks support for parallelization and GPU/TPU execution, making the platform prohibitively slow compared to modern MARL environments and hindering adoption. We contribute JaxRobotarium, a Jax-powered end-to-end simulation, learning, deployment, and benchmarking platform for the Robotarium. JaxRobotarium enables rapid training and deployment of multi-robot RL (MRRL) policies with realistic robot dynamics and safety constraints, supporting parallelization and hardware acceleration. Our generalizable learning interface integrates easily with SOTA MARL libraries (e.g., JaxMARL). In addition, JaxRobotarium includes eight standardized coordination scenarios, including four novel scenarios that bring established MARL benchmark tasks (e.g., RWARE and Level-Based Foraging) to a robotics setting. We demonstrate that JaxRobotarium retains high simulation fidelity while achieving dramatic speedups over baseline (20x in training and 150x in simulation), and provides an open-access sim-to-real evaluation pipeline through the Robotarium testbed, accelerating and democratizing access to multi-robot learning research and evaluation. Our code is available at https://github.com/GT-STAR-Lab/JaxRobotarium.

cs.RO↗

arXiv2025

Segmentation as A Plug-and-Play Capability for Frozen Multimodal LLMs

Integrating diverse visual capabilities into a unified model is a significant trend in Multimodal Large Language Models (MLLMs). Among these, the inclusion of segmentation poses a distinct set of challenges. To equip MLLMs with pixel-level segmentation abilities, prevailing methods require finetuning the model to produce specific outputs compatible with a mask decoder. This process typically alters the model's output space and compromises its intrinsic generalization, which undermines the goal of building a unified model. We introduce LENS (Leveraging kEypoiNts for MLLMs' Segmentation), a novel plug-and-play solution. LENS attaches a lightweight, trainable head to a completely frozen MLLM. By refining the spatial cues embedded in attention maps, LENS extracts keypoints and describes them into point-wise features directly compatible with the mask decoder. Extensive experiments validate our approach: LENS achieves segmentation performance competitive with or superior to that of retraining-based methods. Crucially, it does so while fully preserving the MLLM's generalization capabilities, which are significantly degraded by finetuning approaches. As such, the attachable design of LENS establishes an efficient and powerful paradigm for extending MLLMs, paving the way for truly multi-talented, unified models.

cs.CV↗

arXiv2025

ViSurf: Visual Supervised-and-Reinforcement Fine-Tuning for Large Vision-and-Language Models

Post-training Large Vision-and-Language Models (LVLMs) typically involves Supervised Fine-Tuning (SFT) for knowledge injection or Reinforcement Learning with Verifiable Rewards (RLVR) for performance enhancement. However, SFT often leads to sub-optimal performance, while RLVR remains constrained by the model's internal knowledge base. While a sequential SFT $\rightarrow$ RLVR pipeline can be used, it introduces significant computational overhead and suffers from catastrophic forgetting. To address these limitations, we propose ViSurf (\textbf{Vi}sual \textbf{Su}pervised-and-\textbf{R}einforcement \textbf{F}ine-Tuning), a unified, single-stage paradigm that integrates the strengths of both SFT and RLVR. By analyzing their training objectives, we establish a unified framework that injects ground-truth labels directly into RLVR rollouts, facilitating simultaneous external supervision and internal reinforcement. Furthermore, we introduce three novel reward control strategies to ensure training stability and optimization. Extensive experiments demonstrate that ViSurf consistently outperforms standalone SFT, RLVR, and the traditional two-stage pipeline across diverse benchmarks. In-depth analysis corroborates these findings, validating the derivation and design principles of ViSurf.

cs.CV↗

arXiv2025

Failure-Aware Multi-Robot Coordination for Resilient and Adaptive Target Tracking

Multi-robot coordination is crucial for autonomous systems, yet real-world deployments often encounter various failures. These include both temporary and permanent disruptions in sensing and communication, which can significantly degrade system robustness and performance if not explicitly modeled. Despite its practical importance, failure-aware coordination remains underexplored in the literature. To bridge the gap between idealized conditions and the complexities of real-world environments, we propose a unified failure-aware coordination framework designed to enable resilient and adaptive multi-robot target tracking under both temporary and permanent failure conditions. Our approach systematically distinguishes between two classes of failures: (1) probabilistic and temporary disruptions, where robots recover from intermittent sensing or communication losses by dynamically adapting paths and avoiding inferred danger zones, and (2) permanent failures, where robots lose sensing or communication capabilities irreversibly, requiring sustained, decentralized behavioral adaptation. To handle these scenarios, the robot team is partitioned into subgroups. Robots that remain connected form a communication group and collaboratively plan using partially centralized nonlinear optimization. Robots experiencing permanent disconnection or failure continue to operate independently through decentralized or individual optimization, allowing them to contribute to the task within their local context. We extensively evaluate our method across a range of benchmark variations and conduct a comprehensive assessment under diverse real-world failure scenarios. Results show that our framework consistently achieves robust performance in realistic environments with unknown danger zones, offering a practical and generalizable solution for the multi-robot systems community.

cs.RO↗

arXiv2025

Resilient Multi-Robot Target Tracking with Sensing and Communication Danger Zones

Multi-robot collaboration for target tracking in adversarial environments poses significant challenges, including system failures, dynamic priority shifts, and other unpredictable factors. These challenges become even more pronounced when the environment is unknown. In this paper, we propose a resilient coordination framework for multi-robot, multi-target tracking in environments with unknown sensing and communication danger zones. We consider scenarios where failures caused by these danger zones are probabilistic and temporary, allowing robots to escape from danger zones to minimize the risk of future failures. We formulate this problem as a nonlinear optimization with soft chance constraints, enabling real-time adjustments to robot behaviors based on varying types of dangers and failures. This approach dynamically balances target tracking performance and resilience, adapting to evolving sensing and communication conditions in real-time. To validate the effectiveness of the proposed method, we assess its performance across various tracking scenarios, benchmark it against methods without resilient adaptation and collaboration, and conduct several real-world experiments.

cs.RO↗

arXiv2025

Empowering Small VLMs to Think with Dynamic Memorization and Exploration

Small-scale Vision-Language Models (SVLMs) are exceptionally well-suited for proprietary tasks. Equipping them with thinking capabilities is a critical step to enhance their performance and reliability in these specific domains. However, existing training paradigms, including Supervised Fine-Tuning (SFT) and Reinforcement Learning with Verifiable Reward (RLVR), impose substantial demands on the base VLM, exceeding the capacity of SVLMs. Consequently, directly applying these paradigms to SVLMs fails to instill the desired thinking abilities. A natural solution is to combine SFT and RLVR, leveraging their complementarity to reduce the dependence on model capacity. Yet the core challenge lies in managing the inherent trade-off: excessive reliance on SFT can force the model to memorize pseudo thinking traces, while over-emphasizing RLVR can lead to unstable exploration (i.e., advantage collapse). To address this, we propose DyME, a novel training paradigm that Dynamically selects between Memorization (via SFT) and Exploration (via RLVR) at each optimization step. By ensuring that every update contributes to the trade-off, DyME serves as a robust, standalone strategy that stabilizes SVLM learning. Complementing this paradigm, we further introduce a synergistic Visual Supervision mechanism (comprising a visual checker and refiner) designed to inject dynamically enhanced, image-grounded guidance during optimization. Extensive experiments across diverse domains demonstrate that DyME consistently achieves this balance, and thus delivers substantial performance improvements on specialized tasks. These results establish DyME as a practical and effective solution for empowering SVLMs with reliable thinking capabilities. GitHub: https://github.com/HKUST-LongGroup/DyME

cs.CV↗

arXiv2025

A Comprehensive Survey on Artificial Intelligence for Complex Network: Potential, Methodology and Application

Complex networks pervade various real-world systems, from the natural environment to human societies. The essence of these networks is in their ability to transition and evolve from microscopic disorder-where network topology and node dynamics intertwine-to a macroscopic order characterized by certain collective behaviors. Over the past two decades, complex network science has significantly enhanced our understanding of the statistical mechanics, structures, and dynamics underlying real-world networks. Despite these advancements, there remain considerable challenges in exploring more realistic systems and enhancing practical applications. The emergence of artificial intelligence (AI) technologies, coupled with the abundance of diverse real-world network data, has heralded a new era in complex network science research. This survey aims to systematically address the potential advantages of AI in overcoming the lingering challenges of complex network research. It endeavors to summarize the pivotal research problems and provide an exhaustive review of the corresponding methodologies and applications. Through this comprehensive survey-the first of its kind on AI for complex networks-we expect to provide valuable insights that will drive further research and advancement in this interdisciplinary field.

cs.SI↗

arXiv2025

Beyond Equilibrium: Non-Equilibrium Foundations Should Underpin Generative Processes in Complex Dynamical Systems

This position paper argues that next-generation non-equilibrium-inspired generative models will provide the essential foundation for better modeling real-world complex dynamical systems. While many classical generative algorithms draw inspiration from equilibrium physics, they are fundamentally limited in representing systems with transient, irreversible, or far-from-equilibrium behavior. We show that non-equilibrium frameworks naturally capture non-equilibrium processes and evolving distributions. Through empirical experiments on a dynamic Printz potential system, we demonstrate that non-equilibrium generative models better track temporal evolution and adapt to non-stationary landscapes. We further highlight future directions such as integrating non-equilibrium principles with generative AI to simulate rare events, inferring underlying mechanisms, and representing multi-scale dynamics across scientific domains. Our position is that embracing non-equilibrium physics is not merely beneficial--but necessary--for generative AI to serve as a scientific modeling tool, offering new capabilities for simulating, understanding, and controlling complex systems.

cs.CE↗

arXiv2025

PhD: A ChatGPT-Prompted Visual hallucination Evaluation Dataset

Multimodal Large Language Models (MLLMs) hallucinate, resulting in an emerging topic of visual hallucination evaluation (VHE). This paper contributes a ChatGPT-Prompted visual hallucination evaluation Dataset (PhD) for objective VHE at a large scale. The essence of VHE is to ask an MLLM questions about specific images to assess its susceptibility to hallucination. Depending on what to ask (objects, attributes, sentiment, etc.) and how the questions are asked, we structure PhD along two dimensions, i.e. task and mode. Five visual recognition tasks, ranging from low-level (object / attribute recognition) to middle-level (sentiment / position recognition and counting), are considered. Besides a normal visual QA mode, which we term PhD-base, PhD also asks questions with specious context (PhD-sec) or with incorrect context ({PhD-icc), or with AI-generated counter common sense images (PhD-ccs). We construct PhD by a ChatGPT-assisted semi-automated pipeline, encompassing four pivotal modules: task-specific hallucinatory item (hitem) selection, hitem-embedded question generation, specious / incorrect context generation, and counter-common-sense (CCS) image generation. With over 14k daily images, 750 CCS images and 102k VQA triplets in total, PhD reveals considerable variability in MLLMs' performance across various modes and tasks, offering valuable insights into the nature of hallucination. As such, PhD stands as a potent tool not only for VHE but may also play a significant role in the refinement of MLLMs.

cs.CV↗

arXiv2025

Correlated Growth of Causal Networks

The study of causal structure in complex systems has gained increasing attention, with many recent studies exploring causal networks that capture cause-effect relationships across diverse fields. Despite increasing empirical evidence linking causal structures to network topological correlations, the mechanisms underlying the emergence of these correlations in causal networks remain poorly understood. In this work, we propose a general growth framework for causal networks, incorporating two key types of correlations: causal and dynamic. We analytically demonstrate that degree correlations emerge as a consequence of marginal dependencies on these correlations. Our theoretical predictions align quantitatively with empirical data from four large-scale innovation networks. Our theory not only sheds light on the origins of topological correlations but also provides a general framework for understanding correlated growth across causal systems.

physics.soc-ph↗

arXiv2025

Magnifier Prompt: Tackling Multimodal Hallucination via Extremely Simple Instructions

Hallucinations in multimodal large language models (MLLMs) hinder their practical applications. To address this, we propose a Magnifier Prompt (MagPrompt), a simple yet effective method to tackle hallucinations in MLLMs via extremely simple instructions. MagPrompt is based on the following two key principles, which guide the design of various effective prompts, demonstrating robustness: (1) MLLMs should focus more on the image. (2) When there are conflicts between the image and the model's inner knowledge, MLLMs should prioritize the image. MagPrompt is training-free and can be applied to open-source and closed-source models, such as GPT-4o and Gemini-pro. It performs well across many datasets and its effectiveness is comparable or even better than more complex methods like VCD. Furthermore, our prompt design principles and experimental analyses provide valuable insights into multimodal hallucination.

cs.CL↗