Hai Lin

Xiaomeng Liu, Hai Lin, Ben M. Chen

This paper studies the structural controllability of a class of uncertain switched linear systems, where the parameters of subsystems state matrices are either unknown or zero. The structural controllability is a generalization of the traditional controllability concept for dynamical systems, and purely based on the interconnection relation between the state variables and inputs through non-zero elements in the state matrices. In order to illustrate such a relationship, two kinds of graphic representations of switched linear systems are proposed, based on which graph theory based necessary and sufficient characterizations of the structural controllability for switched linear systems are presented. Finally, the paper concludes with discussions on the results and future work.

Mohamadreza Ahmadi, Bo Wu, Hai Lin et al.

Privacy is an increasing concern in cyber-physical systems that operates over a shared network. In this paper, we propose a method for privacy verification of cyber- physical systems modeled by Markov decision processes (MDPs) and partially-observable Markov decision processes (POMDPs) based on barrier certificates. To this end, we consider an opacity-based notion of privacy, which is characterized by the beliefs in system states. We show that the belief update equations can be represented as discrete-time switched systems, for which we propose a set of conditions for privacy verification in terms of barrier certificates. We further demonstrate that, for MDPs and for POMDPs, privacy verification can be computationally implemented by solving a set of semi-definite programs and sum-of-squares programs, respectively. The method is illustrated by an application to privacy verification of an inventory management system.

Hai Lin, Jinhong Yuan, Wei Yu et al.

The recently proposed orthogonal delay-Doppler division multiplexing (ODDM) modulation, which is a delay-Doppler (DD) domain multi-carrier (DDMC) modulation scheme based on the DD domain orthogonal pulse (DDOP), is studied. We first revisit the linear time-varying (LTV) channel model for the wireless channel, and review the conventional multi-carrier (MC) modulation schemes and their design guidelines for both linear time-invariant (LTI) and LTV channels. We then focus on the representation of the LTV channel in an equivalent sampled DD (ESDD) domain, and propose an impulse-function-based transmission strategy for the ESDD channel. Next, we take an in-depth look into the DDOP and show that it achieves orthogonality with respect to the fine time and frequency resolutions in the ESDD domain thus behaves like an impulse function. This allows us to unveil the unique input-output relation of the resultant ODDM modulation over the ESDD channel. We point out that the conventional MC modulation design guidelines based on the Weyl-Heisenberg (WH) frame theory can be relaxed without compromising its orthogonality or violating the WH frame theory. More specifically, for a practical communication system with bandwidth and duration constraints, MC modulation signals can be designed considering so-called local or sufficient (bi)orthogonality, which refers to the (bi)orthogonality among a WH subset for the MC signal within a specific bandwidth and duration. This novel design guideline could potentially open up opportunities for developing future waveforms required by new applications such as communication systems associated with high delay and/or Doppler shifts, as well as integrated sensing and communications.

Zhiyu Liu, Bo Wu, Jin Dai et al.

Control and communication are often tightly coupled in motion planning of networked mobile robots, due to the fact that robotic motions will affect the overall communication quality, and the quality of service (QoS) of the communication among the robots will in turn affect their coordination performance. In this paper, we propose a control theoretical motion planning framework for a team of networked mobile robots in order to accomplish high-level spatial and temporal motion objectives while optimizing communication QoS. Desired motion specifications are formulated as Signal Temporal Logic (STL), whereas the communication performances to be optimized are captured by recently proposed Spatial Temporal Reach and Escape Logic (STREL) formulas. Both the STL and STREL specifications are encoded as mixed integer linear constraints posed on the system and/or environment state variables of the mobile robot network, where satisfactory control strategies can be computed by exploiting a distributed model predictive control (MPC) approach. To the best of the authors' knowledge, we are the first to study controller synthesis for STREL specifications. A two-layer hierarchical MPC procedure is proposed to efficiently solve the problem, whose soundness and completeness are formally ensured. The effectiveness of the proposed framework is validated by simulation examples.

Yajuan Sun, Hai Lin, Ben M. Chen

This paper investigates the supervisory control of nondeterministic discrete event systems to enforce bisimilarity with respect to deterministic specifications. A notion of synchronous simulation-based controllability is introduced as a necessary and sufficient condition for the existence of a bisimilarity enforcing supervisor, and a polynomial algorithm is developed to verify such a condition. When the existence condition holds, a supervisor achieving bisimulation equivalence is constructed. Furthermore, when the existence condition does not hold, two different methods are provided for synthesizing maximal permissive sub-specifications.

Xiaomeng Liu, Hai Lin, Ben M. Chen

This paper considers the controllability problem for multi-agent systems. In particular, the structural controllability of multi-agent systems under switching topologies is investigated. The structural controllability of multi-agent systems is a generalization of the traditional controllability concept for dynamical systems, and purely based on the communication topologies among agents. The main contributions of the paper are graph-theoretic characterizations of the structural controllability for multi-agent systems. It turns out that the multi-agent system with switching topology is structurally controllable if and only if the union graph G of the underlying communication topologies is connected (single leader) or leader-follower connected (multi-leader). Finally, the paper concludes with several illustrative examples and discussions of the results and future work.

Hai Lin

Large language models are undergoing a transition from model technology to system technology. As developers use Codex, Claude Code, AutoGPT, and related agents to write code, manage projects, and execute multi-step tasks, recurring engineering problems such as cache reuse, context management, agent scheduling, and permission control increasingly resemble classical computer systems problems. This paper develops that analogy as a visionary survey. We map concepts from computer architecture to the emerging model-native stack and review work on LLM-as-OS, memory management, agent frameworks, tool protocols, multi-agent coordination, cognitive architectures, and safety governance. We argue that these strands address different layers of the same system but lack a unified model. To fill this gap, we propose the Intelligent Computing Architecture Model (ICAM), a six-layer framework for model-native computing with explicit interface contracts and design axioms. ICAM resolves the apparent tension over whether an LLM is more like a CPU or an operating system through a dual-plane view: a probabilistic execution plane concerned with what can be computed, and a deterministic control plane concerned with what should be computed. We further introduce three design laws: the Semantic Locality Law for KV-cache reuse and inference speedup, the Context Budget Law for effective working sets under finite windows and attention decay, and the Agent Speedup Law for diminishing returns in multi-agent collaboration. We validate these laws against published system-level data and relate them to recent evidence on agentic software practices. We conclude by identifying where the analogy breaks down and outlining a research roadmap for model-native computing. This is a conceptual and survey contribution; it does not report new experiments.

Mohammad Karimadini, Hai Lin

A natural way for cooperative tasking in multi-agent systems is through a top-down design by decomposing a global task into sub-tasks for each individual agent such that the accomplishments of these sub-tasks will guarantee the achievement of the global task. In our previous works [1], [2] we presented necessary and sufficient conditions on the decomposability of a global task automaton between cooperative agents. As a follow-up work, this paper deals with the robustness issues of the proposed top-down design approach with respect to event failures in the multi-agent systems. The main concern under event failure is whether a previously decomposable task can still be achieved collectively by the agents, and if not, we would like to investigate that under what conditions the global task could be robustly accomplished. This is actually the fault-tolerance issue of the top-down design, and the results provide designers with hints on which events are fragile with respect to failures, and whether redundancies are needed. The main objective of this paper is to identify necessary and sufficient conditions on failed events under which a decomposable global task can still be achieved successfully. For such a purpose, a notion called passivity is introduced to characterize the type of event failures. The passivity is found to reflect the redundancy of communication links over shared events, based on which necessary and sufficient conditions for the reliability of cooperative tasking under event failures are derived, followed by illustrative examples and remarks for the derived conditions.

Mohammad Karimadini, Hai Lin

In our previous work [1], a divide-and-conquer approach was proposed for cooperative tasking among multi-agent systems. The basic idea is to decompose a requested global specification into subtasks for individual agents such that the fulfillment of these subtasks by each individual agent leads to the satisfaction of the global specification as a team. It was shown that not all tasks can be decomposed. Furthermore, a necessary and sufficient condition was proposed for the decomposability of a task automaton between two cooperative agents. The current paper continues the results in [1] and proposes necessary and sufficient conditions for task decomposability with respect to arbitrary finite number of agents. It is further shown that the fulfillment of local specifications can guarantee the satisfaction of the global specification. This work provides hints for the designers on how to rule out the indecomposable task automata and enforce the decomposability conditions. The result therefore may pave the way towards a new perspective for decentralized cooperative control of multi-agent systems.

Xinhua Wang, Hai Lin

In this paper, a continuous finite-time-convergent differentiator is presented based on a strong Lyapunov function. The continuous differentiator can reduce chattering phenomenon sufficiently than normal sliding mode differentiator, and the outputs of signal tracking and derivative estimation are all smooth. Frequency analysis is applied to compare the continuous differentiator with sliding mode differentiator. The beauties of the continuous finite-time-convergent differentiator include its simplicity, restraining noises sufficiently, and avoiding the chattering phenomenon.

Bo Wu, Xiaobin Zhang, Hai Lin

This paper considers the permissive supervisor synthesis for probabilistic systems modeled as Markov Decision Processes (MDP). Such systems are prevalent in power grids, transportation networks, communication networks and robotics. Unlike centralized planning and optimization based planning, we propose a novel supervisor synthesis framework based on learning and compositional model checking to generate permissive local supervisors in a distributed manner. With the recent advance in assume-guarantee reasoning verification for probabilistic systems, building the composed system can be avoided to alleviate the state space explosion and our framework learn the supervisors iteratively based on the counterexamples from verification. Our approach is guaranteed to terminate in finite steps and to be correct.

Xiaobin Zhang, Bo Wu, Hai Lin

Partially Observable Markov Decision Process (POMDP) is widely used to model probabilistic behavior for complex systems. Compared with MDPs, POMDP models a system more accurate but solving a POMDP generally takes exponential time in the size of its state space. This makes the formal verification and synthesis problems much more challenging for POMDPs, especially when multiple system components are involved. As a promising technique to reduce the verification complexity, the abstraction method tries to find an abstract system with a smaller state space but preserves enough properties for the verification purpose. While abstraction based verification has been explored extensively for MDPs, in this paper, we present the first result of POMDP abstraction and its refinement techniques. The main idea follows the counterexample-guided abstraction refinement (CEGAR) framework. Starting with a coarse guess for the POMDP abstraction, we iteratively use counterexamples from formal verification to refine the abstraction until the abstract system can be used to infer the verification result for the original POMDP. Our main contributions have two folds: 1) we propose a novel abstract system model for POMDP and a new simulation relation to capture the partial observability then prove the preservation on a fragment of Probabilistic Computation Tree Logic (PCTL); 2) to find a proper abstract system that can prove or disprove the satisfaction relation on the concrete POMDP, we develop a novel refinement algorithm. Our work leads to a sound and complete CEGAR framework for POMDP.

Quan Quan, Hai Lin, Kai-Yuan Cai

Besides parametric uncertainties and disturbances, the unmodeled dynamics and time delay at the input are often present in practical systems, which cannot be ignored in some cases. This paper aims to solve output feedback tracking control problem for a class of nonlinear uncertain systems subject to unmodeled high-frequency gains and time delay at the input. By the additive decomposition, the uncertain system is transformed to an uncertainty-free system, where the uncertainties, disturbance and effect of unmodeled dynamics plus time delay are lumped into a new disturbance at the output. Sequently, additive decomposition is used to decompose the transformed system, which simplifies the tracking controller design. To demonstrate the effectiveness, the proposed control scheme is applied to three benchmark examples.

Xinhua Wang, Hai Lin

In this paper, a continuous hybrid differentiator is presented based on a strong Lyapunov function. The differentiator design can not only reduce sufficiently chattering phenomenon of derivative estimation by introducing a perturbation parameter, but also the dynamical performances are improved by adding linear correction terms to the nonlinear ones. Moreover, strong robustness ability is obtained by integrating sliding mode items and the linear filter. Frequency analysis is applied to compare the hybrid continuous differentiator with sliding mode differentiator. The merits of the continuous hybrid differentiator include the excellent dynamical performances, restraining noises sufficiently, and avoiding the chattering phenomenon.

Shaobin Cao, Zhijian Ji, Hai Lin et al.

This note investigates how to design topology structures to ensure the controllability of multi-agent networks (MASs) under any selection of leaders. We put forward a concept of perfect controllability, which means that a multi-agent system is controllable with no matter how the leaders are chosen. In this situation, both the number and the locations of leader agents are arbitrary. A necessary and sufficient condition is derived for the perfect controllability. Moreover, a step-by-step design procedure is proposed by which topologies are constructed and are proved to be perfectly controllable. The principle of the proposed design method is interpreted by schematic diagrams along with the corresponding topology structures from simple to complex. We show that the results are valid for any number and any location of leaders. Both the construction process and the corresponding topology structures are clearly outlined.

Mohammad Karimadini, Hai Lin

New advances in large scale distributed systems have amazingly offered complex functionalities through parallelism of simple and rudimentary components. The key issue in cooperative control of multi-agent systems is the synthesis of local control and interaction rules among the agents such that the entire controlled system achieves a desired global behavior. For this purpose, three fundamental problems have to be addressed: (1) task decomposition for top-down design, such that the fulfillment of local tasks guarantees the satisfaction of the global task, by the team; (2) fault-tolerant top-down design, such that the global task remains decomposable and achievable, in spite of some failures, and (3) design of interactions among agents to make an undecomposable task decomposable and achievable in a top-down framework. The first two problems have been addressed in our previous works, by identifying necessary and sufficient conditions for task automaton decomposition, and fault-tolerant task decomposability. This paper deals with the third problem and proposes a procedure to redistribute the events among agents in order to enforce decomposability of an undecomposable task automaton. The decomposability conditions are used to identify the root causes of undecomposability which are found to be due to over-communications that have to be deleted, while respecting the fault-tolerant decomposability conditions; or because of the lack of communications that require new sharing of events, while considering new violations of decomposability conditions. This result provides a sufficient condition to make any undecomposable deterministic task automaton decomposable in order to facilitate cooperative tasking. Illustrative examples are presented to show the concept of task automaton decomposabilization.

Yajuan Sun, Hai Lin, Ben M. Chen

This paper presents an input-output simulation approach to controlling multi-affine systems for linear temporal logic (LTL) specifications, which consists of the following steps. First, we partition the state space into rectangles, each of which satisfies atomic LTL propositions. Then, we study the control of multi-affine systems on rectangles including the control of driving all trajectories starting from a rectangle to exit through a facet and the control of stabilizing the system towards a desired point. With the proposed controllers, a finitely abstracted transition system is constructed which is shown to be input-output simulated by the rectangular transition system of the multi-affine system. Since input-output simulation preserves LTL properties, the controller synthesis of the multi-affine system for LTL specifications is achieved by designing a nonblocking supervisor for the abstracted transition system and by continuously implementing the resulting supervisor for the original multi-affine system.

Yajuan Sun, Hai Lin, Ben. M. Chen

In decentralized systems, branching behaviors naturally arise due to communication, unmodeled dynamics and system abstraction, which can not be adequately captured by the traditional sequencing-based language equivalence. As a finer behavior equivalence than language equivalence, bisimulation not only allows the full set of branching behaviors but also explicitly specifies the properties in terms of temporal logic such as CTL* and mu-calculus. This observation motivates us to consider the decentralized control of discrete event systems (DESs) for bisimulation equivalence in this paper, where the plant and the specification are taken to be nondeterministic and the supervisor is taken to be deterministic. An automata-based control framework is formalized, upon which we develop three architectures with respect to different decision fusion rules for the decentralized bisimilarity control, named a conjunctive architecture, a disjunctive architecture and a general architecture. Under theses three architectures, necessary and sufficient conditions for the existence of decentralized bisimilarity supervisors are derived respectively, which extend the traditional results of supervisory control from language equivalence to bisimulation equivalence. It is shown that these conditions can be verified with exponential complexity. Furthermore, the synthesis of bisimilarity supervisors is presented when the existence condition holds.

Jin Dai, Hai Lin

We investigate the coordination and control problems of distributed discrete event systems that are composed of multiple subsystems subject to potential actuator and/or sensor faults. We model actuator faults as local controllability loss of certain actuator events and sensor faults as observability failure of certain sensor readings, respectively. Starting from automata-theoretic models that characterize behaviors of the subsystems in the presence of faulty actuators and/or sensors, we establish necessary and sufficient conditions for the existence of actuator and sensor fault tolerant supervisors, respectively, and synthesize appropriate local post-fault supervisors to prevent the post-fault subsystems from jeopardizing local safety requirements. Furthermore, we apply an assume-guarantee coordination scheme to the controlled subsystems for both the nominal and faulty subsystems so as to achieve the desired specifications of the system. A multi-robot coordination example is used to illustrate the proposed coordination and control architecture.

Hai Lin, Shaoxiong Zhan, Junyou Su et al.

In Retrieval-Augmented Generation (RAG) tasks using Large Language Models (LLMs), the quality of retrieved information is critical to the final output. This paper introduces the IRSC benchmark for evaluating the performance of embedding models in multilingual RAG tasks. The benchmark encompasses five retrieval tasks: query retrieval, title retrieval, part-of-paragraph retrieval, keyword retrieval, and summary retrieval. Our research addresses the current lack of comprehensive testing and effective comparison methods for embedding models in RAG scenarios. We introduced new metrics: the Similarity of Semantic Comprehension Index (SSCI) and the Retrieval Capability Contest Index (RCCI), and evaluated models such as Snowflake-Arctic, BGE, GTE, and M3E. Our contributions include: 1) the IRSC benchmark, 2) the SSCI and RCCI metrics, and 3) insights into the cross-lingual limitations of embedding models. The IRSC benchmark aims to enhance the understanding and development of accurate retrieval systems in RAG tasks. All code and datasets are available at: https://github.com/Jasaxion/IRSC_Benchmark

Yajuan Sun, Hai Lin, Fuchun Liu

Bisimulation relation has been successfully applied to computer science and control theory. In our previous work, simulation-based controllability and simulation-based observability are proposed, under which the existence of bisimilarity supervisor is guaranteed. However, a given specification automaton may not satisfy these conditions, and a natural question is how to compute a maximum permissive subspecification. This paper aims to answer this question and investigate the computation of the supremal simulation-based controllable and strong observable subautomata with respect to given specifications by the lattice theory. In order to achieve the supremal solution, three monotone operators, namely simulation operator, controllable operator and strong observable operator, are proposed upon the established complete lattice. Then, inequalities based on these operators are formulated, whose solution is the simulation-based controllable and strong observable set. In particular, a sufficient condition is presented to guarantee the existence of the supremal simulation-based controllable and strong observable subautomata. Furthermore, an algorithm is proposed to compute such subautomata.

Xiaodong Cai, Hai Lin, Shaoxiong Zhan et al.

Token sampling strategies critically influence text generation quality in large language models (LLMs). However, existing methods introduce additional hyperparameters, requiring extensive tuning and complicating deployment. We present Entropy Equilibrium Sampling (EES), an auxiliary hyperparameter-free approach inspired by information theory that can dynamically adjust candidate sets by balancing normalized entropy with probability mass. We evaluate EES on both reasoning and generation tasks across a range of model architectures. Our results show that EES consistently performs well across temperature settings, delivering competitive accuracy and coherence while maintaining diversity. By eliminating the need for hyperparameter tuning, EES greatly simplifies deployment while improving performance. Code is available at https://github.com/shuanncai/EES

Xiangyang He, Yubo Tao, Shuoliu Yang et al.

Relationships in scientific data, such as the numerical and spatial distribution relations of features in univariate data, the scalar-value combinations' relations in multivariate data, and the association of volumes in time-varying and ensemble data, are intricate and complex. This paper presents voxel2vec, a novel unsupervised representation learning model, which is used to learn distributed representations of scalar values/scalar-value combinations in a low-dimensional vector space. Its basic assumption is that if two scalar values/scalar-value combinations have similar contexts, they usually have high similarity in terms of features. By representing scalar values/scalar-value combinations as symbols, voxel2vec learns the similarity between them in the context of spatial distribution and then allows us to explore the overall association between volumes by transfer prediction. We demonstrate the usefulness and effectiveness of voxel2vec by comparing it with the isosurface similarity map of univariate data and applying the learned distributed representations to feature classification for multivariate data and to association analysis for time-varying and ensemble data.

Yuan Li, Huan Liu, Yubo Tao et al.

Recently, deep learning-based tooth segmentation methods have been limited by the expensive and time-consuming processes of data collection and labeling. Achieving high-precision segmentation with limited datasets is critical. A viable solution to this entails fine-tuning pre-trained multiview-based models, thereby enhancing performance with limited data. However, relying solely on two-dimensional (2D) images for three-dimensional (3D) tooth segmentation can produce suboptimal outcomes because of occlusion and deformation, i.e., incomplete and distorted shape perception. To improve this fine-tuning-based solution, this paper advocates 2D-3D joint perception. The fundamental challenge in employing 2D-3D joint perception with limited data is that the 3D-related inputs and modules must follow a lightweight policy instead of using huge 3D data and parameter-rich modules that require extensive training data. Following this lightweight policy, this paper selects skeletons as the 3D inputs and introduces MSFormer, a novel method for tooth segmentation. MSFormer incorporates two lightweight modules into existing multiview-based models: a 3D-skeleton perception module to extract 3D perception from skeletons and a skeleton-image contrastive learning module to obtain the 2D-3D joint perception by fusing both multiview and skeleton perceptions. The experimental results reveal that MSFormer paired with large pre-trained multiview models achieves state-of-the-art performance, requiring only 100 training meshes. Furthermore, the segmentation accuracy is improved by 2.4%-5.5% with the increasing volume of training data.

Xinhua Wang, Hai Lin

This paper presents a high-order differentiator for delayed measurement signal. The proposed differentiator not only can correct the delay in signal, but aslo can estimate the undelayed derivatives. The differentiator consists of two-step algorithms with the delayed time instant. Conditions are given ensuring convergence of the estimation error for the given delay in the signals. The merits of method include its simple implementation and interesting application. Numerical simulations illustrate the effectiveness of the proposed differentiator.

Zhengmeng Xu, Yujie Wang, Xiaotong Feng et al.

We propose a time series forecasting method named Quantum Gramian Angular Field (QGAF). This approach merges the advantages of quantum computing technology with deep learning, aiming to enhance the precision of time series classification and forecasting. We successfully transformed stock return time series data into two-dimensional images suitable for Convolutional Neural Network (CNN) training by designing specific quantum circuits. Distinct from the classical Gramian Angular Field (GAF) approach, QGAF's uniqueness lies in eliminating the need for data normalization and inverse cosine calculations, simplifying the transformation process from time series data to two-dimensional images. To validate the effectiveness of this method, we conducted experiments on datasets from three major stock markets: the China A-share market, the Hong Kong stock market, and the US stock market. Experimental results revealed that compared to the classical GAF method, the QGAF approach significantly improved time series prediction accuracy, reducing prediction errors by an average of 25% for Mean Absolute Error (MAE) and 48% for Mean Squared Error (MSE). This research confirms the potential and promising prospects of integrating quantum computing with deep learning techniques in financial time series forecasting.

Yating Jiang, Hai Lin, Yi-Han Chiang et al.

In this paper, a unified framework for chirp-domain waveforms, including orthogonal chirp division multiplexing (OCDM) and affine frequency division multiplexing (AFDM), is developed. Based on their continuous-time representations, we show that these waveforms fall within the conventional Weyl-Heisenberg (WH) framework for multicarrier (MC) waveforms, where the root chirp corresponds directly to the prototype pulse in the WH framework. Since the chirp is a constant-envelope signal and is transparent to subcarrier orthogonality, these waveforms can be further interpreted as pulse-shaped (PS) orthogonal frequency division multiplexing (OFDM). Within the developed PS-OFDM framework, the power spectral density of chirp-domain waveforms is derived analytically. We then discuss existing practical implementations of chirp-domain waveforms, which rely on sub-Nyquist discrete-time samples and therefore exhibit frequency aliasing. The resulting aliased waveform is analyzed, and the orthogonality among the embedded aliased chirps is discussed. It is shown that the aliased chirps are conditionally orthogonal, whereas the implemented approximate aliased chirps can maintain mutual orthogonality when an appropriate sample-wise pulse-shaping filter is applied. We further derive an exact input-output relation for the implemented chirp-domain waveform over a delay-Doppler (DD) channel, showing that the effective channel observed at a practical receiver does not, in general, admit a DD spreading-function model commonly assumed in the literature. The implementation complexity is also investigated and compared with that of orthogonal delay-Doppler division multiplexing (ODDM), the DD-domain MC waveform defined within the evolved WH framework. Finally, simulation results are provided to verify the analysis.

Shen Han, Zhiyao Zhou, Jiawei Chen et al.

The quality of graph-structured data is fundamental to the success of modern graph analysis techniques such as Graph Neural Networks (GNNs). However, real-world graph data is often suboptimal, suffering from issues such as noise and incomplete connections. Graph Structure Learning (GSL) has emerged as a promising technique that adaptively optimizes node connections. However, we observe that the effectiveness of GSL often comes at the cost of a dramatic expansion in edge count, resulting in significant storage and computational overhead. In this work, we reveal that this limitation stems from the prevalent use of similarity-based edge construction, which predominantly connects highly similar neighbors based on their embeddings, introducing substantial structure redundancy. To address this, we propose a novel Informative Graph Structure Learning method (InGSL), which jointly considers both similarity and diversity in edge construction by incorporating a mutual-information-guided learning strategy. Notably, InGSL serves as a plug-in module that can be seamlessly integrated into existing GSL frameworks. Through extensive experiments on six representative GSL methods, we demonstrate that InGSL achieves significant performance improvements at a reduced number of edges.

Kevin Smith, Hai Lin, Praveen Tiwari et al.

Property Specification Language (PSL) is a form of temporal logic that has been mainly used in discrete domains (e.g. formal hardware verification). In this paper, we show that by merging machine learning techniques with PSL monitors, we can extend PSL to work on continuous domains. We apply this technique in machine learning-based anomaly detection to analyze scenarios of real-time streaming events from continuous variables in order to detect abnormal behaviors of a system. By using machine learning with formal models, we leverage the strengths of both machine learning methods and formal semantics of time. On one hand, machine learning techniques can produce distributions on continuous variables, where abnormalities can be captured as deviations from the distributions. On the other hand, formal methods can characterize discrete temporal behaviors and relations that cannot be easily learned by machine learning techniques. Interestingly, the anomalies detected by machine learning and the underlying time representation used are discrete events. We implemented a temporal monitoring package (TEF) that operates in conjunction with normal data science packages for anomaly detection machine learning systems, and we show that TEF can be used to perform accurate interpretation of temporal correlation between events.

Hongming Tan, Shaoxiong Zhan, Hai Lin et al.

In dense retrieval, embedding long texts into dense vectors can result in information loss, leading to inaccurate query-text matching. Additionally, low-quality texts with excessive noise or sparse key information are unlikely to align well with relevant queries. Recent studies mainly focus on improving the sentence embedding model or retrieval process. In this work, we introduce a novel text augmentation framework for dense retrieval. This framework transforms raw documents into information-dense text formats, which supplement the original texts to effectively address the aforementioned issues without modifying embedding or retrieval methodologies. Two text representations are generated via large language models (LLMs) zero-shot prompting: question-answer pairs and element-driven events. We term this approach QAEA-DR: unifying question-answer generation and event extraction in a text augmentation framework for dense retrieval. To further enhance the quality of generated texts, a scoring-based evaluation and regeneration mechanism is introduced in LLM prompting. Our QAEA-DR model has a positive impact on dense retrieval, supported by both theoretical analysis and empirical experiments.

Yankai Jiang, Peng Zhang, Donglin Yang et al.

We explore Generalizable Tumor Segmentation, aiming to train a single model for zero-shot tumor segmentation across diverse anatomical regions. Existing methods face limitations related to segmentation quality, scalability, and the range of applicable imaging modalities. In this paper, we uncover the potential of the internal representations within frozen medical foundation diffusion models as highly efficient zero-shot learners for tumor segmentation by introducing a novel framework named DiffuGTS. DiffuGTS creates anomaly-aware open-vocabulary attention maps based on text prompts to enable generalizable anomaly segmentation without being restricted by a predefined training category list. To further improve and refine anomaly segmentation masks, DiffuGTS leverages the diffusion model, transforming pathological regions into high-quality pseudo-healthy counterparts through latent space inpainting, and applies a novel pixel-level and feature-level residual learning approach, resulting in segmentation masks with significantly enhanced quality and generalization. Comprehensive experiments on four datasets and seven tumor categories demonstrate the superior performance of our method, surpassing current state-of-the-art models across multiple zero-shot settings. Codes are available at https://github.com/Yankai96/DiffuGTS.

Vince Kurtz, Hai Lin

Contact-implicit trajectory optimization offers an appealing method of automatically generating complex and contact-rich behaviors for robot manipulation and locomotion. The scalability of such techniques has been limited, however, by the challenge of ensuring both numerical reliability and physical realism. In this paper, we present preliminary results suggesting that the Iterative Linear Quadratic Regulator (iLQR) algorithm together with the recently proposed pressure-field-based hydroelastic contact model enables reliable and physically realistic trajectory optimization through contact. We use this approach to synthesize contact-rich behaviors like quadruped locomotion and whole-arm manipulation. Furthermore, open-loop playback on a Kinova Gen3 robot arm demonstrates the physical accuracy of the whole-arm manipulation trajectories. Code is available at https://bit.ly/ilqr_hc and videos can be found at https://youtu.be/IqxJKbM8_ms.

Shunlong Wu, Hai Lin, Shaoshen Chen et al.

Existing KV cache compression methods generally operate on discrete tokens or non-semantic chunks. However, such approaches often lead to semantic fragmentation, where linguistically coherent units are disrupted, causing irreversible information loss and degradation in model performance. To address this, we introduce SemantiCache, a novel compression framework that preserves semantic integrity by aligning the compression process with the semantic hierarchical nature of language. Specifically, we first partition the cache into semantically coherent chunks by delimiters, which are natural semantic boundaries. Within each chunk, we introduce a computationally efficient Greedy Seed-Based Clustering (GSC) algorithm to group tokens into semantic clusters. These clusters are further merged into semantic cores, enhanced by a Proportional Attention mechanism that rebalances the reduced attention contributions of the merged tokens. Extensive experiments across diverse benchmarks and models demonstrate that SemantiCache accelerates the decoding stage of inference by up to 2.61 times and substantially reduces memory footprint, while maintaining performance comparable to the original model.

Yuxuan Kuang, Hai Lin, Meng Jiang · pku

Object navigation (ObjectNav) requires an agent to navigate through unseen environments to find queried objects. Many previous methods attempted to solve this task by relying on supervised or reinforcement learning, where they are trained on limited household datasets with close-set objects. However, two key challenges are unsolved: understanding free-form natural language instructions that demand open-set objects, and generalizing to new environments in a zero-shot manner. Aiming to solve the two challenges, in this paper, we propose OpenFMNav, an Open-set Foundation Model based framework for zero-shot object Navigation. We first unleash the reasoning abilities of large language models (LLMs) to extract proposed objects from natural language instructions that meet the user's demand. We then leverage the generalizability of large vision language models (VLMs) to actively discover and detect candidate objects from the scene, building a Versatile Semantic Score Map (VSSM). Then, by conducting common sense reasoning on VSSM, our method can perform effective language-guided exploration and exploitation of the scene and finally reach the goal. By leveraging the reasoning and generalizing abilities of foundation models, our method can understand free-form human instructions and perform effective open-set zero-shot navigation in diverse environments. Extensive experiments on the HM3D ObjectNav benchmark show that our method surpasses all the strong baselines on all metrics, proving our method's effectiveness. Furthermore, we perform real robot demonstrations to validate our method's open-set-ness and generalizability to real-world environments.

Bukunmi G. Odunlami, Marcos Netto, Hai Lin

Discrete events alter how parameter influence propagates in hybrid systems. Prevailing Fisher information formulations assume that sensitivities evolve smoothly according to continuous-time variational equations and therefore neglect the sensitivity updates induced by discrete events. This paper derives a Fisher information matrix formulation compatible with hybrid systems. To do so, we use the saltation matrix, which encodes the first order transformation of sensitivities induced by discrete events. The resulting formulation is referred to as the salted Fisher information matrix (SFIM). The proposed framework unifies continuous information accumulation during flows with discrete updates at event times. We further establish that hybrid persistence of excitation provides a sufficient condition for positive definiteness of the SFIM. Examples are provided to demonstrate the merit of the proposed approach, including a three bus generator wind turbine differential algebraic power system

Shaoxiong Zhan, Yanlin Lai, Zheng Liu et al.

Current Large Language Models have achieved Olympiad-level logic, yet Vision-Language Models paradoxically falter on elementary spatial tasks like block counting. This capability mismatch reveals a critical ``spatial intelligence gap,'' where models fail to construct coherent 3D mental representations from 2D observations. We uncover this gap via diagnostic analyses showing the bottleneck is a missing view-consistent spatial interface rather than insufficient visual features or weak reasoning. To bridge this, we introduce \textbf{3ViewSense}, a framework that grounds spatial reasoning in Orthographic Views. Drawing on engineering cognition, we propose a ``Simulate-and-Reason'' mechanism that decomposes complex scenes into canonical orthographic projections to resolve geometric ambiguities. By aligning egocentric perceptions with these allocentric references, our method facilitates explicit mental rotation and reconstruction. Empirical results on spatial reasoning benchmarks demonstrate that our method significantly outperforms existing baselines, with consistent gains on occlusion-heavy counting and view-consistent spatial reasoning. The framework also improves the stability and consistency of spatial descriptions, offering a scalable path toward stronger spatial intelligence in multimodal systems.

Fahimeh Orvati Nia, Hai Lin

Accurate pedestrian intention estimation is crucial for the safe navigation of autonomous vehicles (AVs) and hence attracts a lot of research attention. However, current models often fail to adequately consider dynamic traffic signals and contextual scene information, which are critical for real-world applications. This paper presents a Traffic-Aware Spatio-Temporal Graph Convolutional Network (TA-STGCN) that integrates traffic signs and their states (Red, Yellow, Green) into pedestrian intention prediction. Our approach introduces the integration of dynamic traffic signal states and bounding box size as key features, allowing the model to capture both spatial and temporal dependencies in complex urban environments. The model surpasses existing methods in accuracy. Specifically, TA-STGCN achieves a 4.75% higher accuracy compared to the baseline model on the PIE dataset, demonstrating its effectiveness in improving pedestrian intention prediction.

Ning Yang, Hai Lin, Yibo Liu et al.

Ensuring that large language models (LLMs) generate outputs aligned with human preferences is important for safe and effective AI interactions. While Direct Preference Optimization (DPO) employs an implicit reward function to optimize the policy model, however, it and its related variants overlook the differential importance of individual tokens and are sensitive to judgment noise in preference datasets during generation. Although recent methods attempt to assess the important weight of tokens via probability prediction or simplistic weighting schemes, these evaluation methods are prone to biases and still cannot fully address these issues. To solve this problem, we propose the Token-Importance Guided Direct Preference Optimization (TI-DPO), which introduces two key innovations: the gradient-based token-importance weights that dynamically prioritize critical tokens, and a triple loss that explicitly guides model outputs to approach human-preferred responses and stay away from non-preferred responses. Experimental results show that TI-DPO achieves higher accuracy and stronger generative diversity, providing more stable and computationally efficient solutions compared with DPO and other RLHF methods.

Zihao Song, Shirantha Welikala, Panos J. Antsaklis et al.

In this paper, we consider the distributed optimal control problem for discrete-time linear networked systems. In particular, we are interested in learning distributed optimal controllers using graph recurrent neural networks (GRNNs). Most of the existing approaches result in centralized optimal controllers with offline training processes. However, as the increasing demand of network resilience, the optimal controllers are further expected to be distributed, and are desirable to be trained in an online distributed fashion, which are also the main contributions of our work. To solve this problem, we first propose a GRNN-based distributed optimal control method, and we cast the problem as a self-supervised learning problem. Then, the distributed online training is achieved via distributed gradient computation, and inspired by the (consensus-based) distributed optimization idea, a distributed online training optimizer is designed. Furthermore, the local closed-loop stability of the linear networked system under our proposed GRNN-based controller is provided by assuming that the nonlinear activation function of the GRNN-based controller is both local sector-bounded and slope-restricted. The effectiveness of our proposed method is illustrated by numerical simulations using a specifically developed simulator.

Hai Lin, Cheng Huang, Zhihong Chen

Reinforcement learning tasks in real-world scenarios often involve large, high-dimensional action spaces, leading to challenges such as convergence difficulties, instability, and high computational complexity. It is widely acknowledged that traditional value-based reinforcement learning algorithms struggle to address these issues effectively. A prevalent approach involves generating independent sub-actions within each dimension of the action space. However, this method introduces bias, hindering the learning of optimal policies. In this paper, we propose an advantage-based optimization method and an algorithm named Advantage Branching Dueling Q-network (ABQ). ABQ incorporates a baseline mechanism to tune the action value of each dimension, leveraging the advantage relationship across different sub-actions. With this approach, the learned policy can be optimized for each dimension. Empirical results demonstrate that ABQ outperforms BDQ, achieving 3%, 171%, and 84% more cumulative rewards in HalfCheetah, Ant, and Humanoid environments, respectively. Furthermore, ABQ exhibits competitive performance when compared against two continuous action benchmark algorithms, DDPG and TD3.

Federico Ottomano, John Y. Goulermas, Vladimir Gusev et al.

Machine Learning (ML) has offered innovative perspectives for accelerating the discovery of new functional materials, leveraging the increasing availability of material databases. Despite the promising advances, data-driven methods face constraints imposed by the quantity and quality of available data. Moreover, ML is often employed in tandem with simulated datasets originating from density functional theory (DFT), and assessed through in-sample evaluation schemes. This scenario raises questions about the practical utility of ML in uncovering new and significant material classes for industrial applications. Here, we propose a data-driven framework aimed at accelerating the discovery of new transparent conducting materials (TCMs), an important category of semiconductors with a wide range of applications. To mitigate the shortage of available data, we create and validate unique experimental databases, comprising several examples of existing TCMs. We assess state-of-the-art (SOTA) ML models for property prediction from the stoichiometry alone. We propose a bespoke evaluation scheme to provide empirical evidence on the ability of ML to uncover new, previously unseen materials of interest. We test our approach on a list of 55 compositions containing typical elements of known TCMs. Although our study indicates that ML tends to identify new TCMs compositionally similar to those in the training data, we empirically demonstrate that it can highlight material candidates that may have been previously overlooked, offering a systematic approach to identify materials that are likely to display TCMs characteristics.

Yun Liang, Hai Lin, Shaojian Qiu et al.

Recently, Transformers have been introduced into the field of acoustics recognition. They are pre-trained on large-scale datasets using methods such as supervised learning and semi-supervised learning, demonstrating robust generality--It fine-tunes easily to downstream tasks and shows more robust performance. However, the predominant fine-tuning method currently used is still full fine-tuning, which involves updating all parameters during training. This not only incurs significant memory usage and time costs but also compromises the model's generality. Other fine-tuning methods either struggle to address this issue or fail to achieve matching performance. Therefore, we conducted a comprehensive analysis of existing fine-tuning methods and proposed an efficient fine-tuning approach based on Adapter tuning, namely AAT. The core idea is to freeze the audio Transformer model and insert extra learnable Adapters, efficiently acquiring downstream task knowledge without compromising the model's original generality. Extensive experiments have shown that our method achieves performance comparable to or even superior to full fine-tuning while optimizing only 7.118% of the parameters. It also demonstrates superiority over other fine-tuning methods.

Vince Kurtz, Patrick M. Wensing, Hai Lin

Task-space Passivity-Based Control (PBC) for manipulation has numerous appealing properties, including robustness to modeling error and safety for human-robot interaction. Existing methods perform poorly in singular configurations, however, such as when all the robot's joints are fully extended. Additionally, standard methods for constrained task-space PBC guarantee passivity only when constraints are not active. We propose a convex-optimization-based control scheme that provides guarantees of singularity avoidance, passivity, and feasibility. This work paves the way for PBC with passivity guarantees under other types of constraints as well, including joint limits and contact/friction constraints. The proposed methods are validated in simulation experiments on a 7 degree-of-freedom manipulator.

Vince Kurtz, He Li, Patrick M. Wensing et al.

Seemingly in defiance of basic physics, cats consistently land on their feet after falling. In this paper, we design a controller that lands the Mini Cheetah quadruped robot on its feet as well. Specifically, we explore how trajectory optimization and machine learning can work together to enable highly dynamic bioinspired behaviors. We find that a reflex approach, in which a neural network learns entire state trajectories, outperforms a policy approach, in which a neural network learns a mapping from states to control inputs. We validate our proposed controller in both simulation and hardware experiments, and are able to land the robot on its feet from falls with initial pitch angles between -90 and 90 degrees.

Yuan Li, Chuanchang Chen, Yubo Tao et al.

This paper proposes a novel model for predicting subgraphs in dynamic graphs, an extension of traditional link prediction. This proposed end-to-end model learns a mapping from the subgraph structures in the current snapshot to the subgraph structures in the next snapshot directly, i.e., edge existence among multiple nodes in the subgraph. A new mechanism named cross-attention with a twin-tower module is designed to integrate node attribute information and topology information collaboratively for learning subgraph evolution. We compare our model with several state-of-the-art methods for subgraph prediction and subgraph pattern prediction in multiple real-world homogeneous and heterogeneous dynamic graphs, respectively. Experimental results demonstrate that our model outperforms other models in these two tasks, with a gain increase from 5.02% to 10.88%.

Tongjia Zheng, Hai Lin

Recent years have seen an increased interest in using mean-field density based modelling and control strategy for deploying robotic swarms. In this paper, we study how to dynamically deploy the robots subject to their physical constraints to efficiently measure and reconstruct certain unknown spatial field (e.g. the air pollution index over a city). Specifically, the evolution of the robots' density is modelled by mean-field partial differential equations (PDEs) which are uniquely determined by the robots' individual dynamics. Bayesian regression models are used to obtain predictions and return a variance function that represents the confidence of the prediction. We formulate a PDE constrained optimization problem based on this variance function to dynamically generate a reference density signal which guides the robots to uncertain areas to collect new data, and design mean-field feedback-based control laws such that the robots' density converges to this reference signal. We also show that the proposed feedback law is robust to density estimation errors in the sense of input-to-state stability. Simulations are included to verify the effectiveness of the algorithms.

Vince Kurtz, Hai Lin

Signal Temporal Logic (STL) has gained popularity in recent years as a specification language for cyber-physical systems, especially in robotics. Beyond being expressive and easy to understand, STL is appealing because the synthesis problem---generating a trajectory that satisfies a given specification---can be formulated as a trajectory optimization problem. Unfortunately, the associated cost function is nonsmooth and non-convex. As a result, existing synthesis methods scale poorly to high-dimensional nonlinear systems. In this letter, we present a new trajectory optimization approach for STL synthesis based on Differential Dynamic Programming (DDP). It is well known that DDP scales well to extremely high-dimensional nonlinear systems like robotic quadrupeds and humanoids: we show that these advantages can be harnessed for STL synthesis. We prove the soundness of our proposed approach, demonstrate order-of-magnitude speed improvements over the state-of-the-art on several benchmark problems, and demonstrate the scalability of our approach to the full nonlinear dynamics of a 7 degree-of-freedom robot arm.

Rafael Rodrigues da Silva, Vince Kurtz, Hai Lin

Many safety-critical systems must achieve high-level task specifications with guaranteed safety and correctness. Much recent progress towards this goal has been made through controller synthesis from temporal logic specifications. Existing approaches, however, have been limited to relatively short and simple specifications. Furthermore, existing methods either consider some prior discretization of the state-space, deal only with a convex fragment of temporal logic, or are not provably complete. We propose a scalable, provably complete algorithm that synthesizes continuous trajectories to satisfy non-convex \gls*{rtl} specifications. We separate discrete task planning and continuous motion planning on-the-fly and harness highly efficient boolean satisfiability (SAT) and \gls*{lp} solvers to find dynamically feasible trajectories that satisfy non-convex \gls*{rtl} specifications for high dimensional systems. The proposed design algorithms are proven sound and complete, and simulation results demonstrate our approach's scalability.

Zhiyu Liu, Meng Jiang, Hai Lin

We aim to enable an autonomous robot to learn new skills from demo videos and use these newly learned skills to accomplish non-trivial high-level tasks. The goal of developing such autonomous robot involves knowledge representation, specification mining, and automated task planning. For knowledge representation, we use a graph-based spatial temporal logic (GSTL) to capture spatial and temporal information of related skills demonstrated by demo videos. We design a specification mining algorithm to generate a set of parametric GSTL formulas from demo videos by inductively constructing spatial terms and temporal formulas. The resulting parametric GSTL formulas from specification mining serve as a domain theory, which is used in automated task planning for autonomous robots. We propose an automatic task planning based on GSTL where a proposer is used to generate ordered actions, and a verifier is used to generate executable task plans. A table setting example is used throughout the paper to illustrate the main ideas.

Tongjia Zheng, Qing Han, Hai Lin

With the rapid development of AI and robotics, transporting a large swarm of networked robots has foreseeable applications in the near future. Existing research in swarm robotics has mainly followed a bottom-up philosophy with predefined local coordination and control rules. However, it is arduous to verify the global requirements and analyze their performance. This motivates us to pursue a top-down approach, and develop a provable control strategy for deploying a robotic swarm to achieve a desired global configuration. Specifically, we use mean-field partial differential equations (PDEs) to model the swarm and control its mean-field density (i.e., probability density) over a bounded spatial domain using mean-field feedback. The presented control law uses density estimates as feedback signals and generates corresponding velocity fields that, by acting locally on individual robots, guide their global distribution to a target profile. The design of the velocity field is therefore centralized, but the implementation of the controller can be fully distributed -- individual robots sense the velocity field and derive their own velocity control signals accordingly. The key contribution lies in applying the concept of input-to-state stability (ISS) to show that the perturbed closed-loop system (a nonlinear and time-varying PDE) is locally ISS with respect to density estimation errors. The effectiveness of the proposed control laws is verified using agent-based simulations.