arXiv:2605.19678v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have shown strong performance on embodied manipulation, yet they remain brittle under visual observation changes, paraphrased language instructions, and compounded perturbations. This limitation suggests that existing methods still rely heavily on shallow correlations in the training distribution, rather than learning stable couplings among task semantics, environment states, and action generation. Although recent efforts improve robustness through larger-scale training, post-training adaptation, or enhanced predictive modeling, they rarely enforce invariance-oriented consistency within the end-to-end policy itself. To address this issue, we propose RoVLA, a robust vision-language-action framework with multi-consistency constraints. RoVLA enforces consistency under three complementary transformations: instruction semantics, trajectory evolution, and observation perturbation. Specifically, Instructional Consistency (IC) promotes stable grounding under semantically equivalent instruction rewrites, Evolutionary Consistency (EC) preserves coherent action intent throughout the generation process, and Observational Consistency (OC) improves robustness to visual and proprioceptive perturbations by enforcing consistent predictions before and after targeted disturbances. By explicitly modeling these invariances during training, RoVLA reduces reliance on superficial correlations and improves robustness and generalization. Experiments on LIBERO-Plus, RoboTwin 2.0, and real-world manipulation tasks show that RoVLA consistently outperforms strong baseline methods and exhibits superior robustness under diverse task and observation shifts. These results demonstrate the effectiveness of multi-consistency learning for robust embodied control. Codes will be available at https://github.com/HCPLab-SYSU/RoVLA.
Science Journals
arXiv:2503.06310v4 Announce Type: replace Abstract: Generating coherent long-form video sequences from discrete text prompts remains challenging due to difficulties in maintaining temporal coherence, semantic consistency, and scene-action continuity across segments. We propose a novel storytelling framework that integrates scene and action prompts through dynamics-inspired prompt mixing. Our approach combines three key components: (i) a bidirectional time-weighted latent blending strategy that enforces temporal consistency between consecutive video segments, (ii) a dynamics-informed prompt weighting (DIPW) mechanism that adaptively balances scene and action prompts at each diffusion timestep based on CLIP-based alignment, narrative progression, and temporal smoothness, and (iii) a semantic action representation that encodes high-level action semantics to modulate transitions according to action similarity. Latent-space blending preserves spatial coherence within scenes, while time-weighted blending introduces bidirectional temporal constraints to prevent abrupt transitions. Together, these components enable fluid and coherent video narratives that faithfully reflect both scene context and action dynamics. Extensive experiments demonstrate that our method significantly outperforms baselines, producing temporally consistent and visually compelling long-form videos without any additional training, thereby bridging the gap between short clips and extended text-driven video storytelling.
arXiv:2605.19454v1 Announce Type: new Abstract: We derive a primal discontinuous Galerkin (DG) formulation for heterogeneous and anisotropic diffusion, obtained by exact algebraic elimination of the skeletal unknown in a compact hybridized interior penalty (H-IP) method. The resulting Unified Interior Penalty DG (UIP-DG) scheme involves transmissibility-based weights inherited from the hybrid formulation, together with two stabilization terms acting respectively on the primal jump and on the jump of the normal diffusive flux. These penalties scale, respectively, with the harmonic mean and with the inverse arithmetic mean of the face-wise transmissibilities. This construction provides a unified perspective on several interior penalty approaches previously introduced independently, while yielding a robust method with stability properties independent of the diffusion contrast and anisotropy. We prove consistency, coercivity, and boundedness of the formulation, and derive quasi-optimal energy-norm a priori error estimates for all variants. Numerical experiments confirm the theoretical claims.
arXiv:2605.19951v1 Announce Type: new Abstract: We present a novel parallel implementation for large-scale three-dimensional electromagnetic inversion based on a Gauss-Newton framework combined with a rational near-best approximation of the matrix exponential for transient simulations. The method employs parallel direct solvers for the shifted linear systems arising from the partial fraction representation of the rational approximation and demonstrates efficient parallel execution on a shared-memory architecture using MPI. A key property of the approach is that the time dependence is entirely contained in the residuals of the employed rational functions, such that the computation of forward responses and sensitivities becomes effectively independent of the number of desired observation times. Model regularization is done with smoothness constraints, formulated with Raviart-Thomas elements. The linearized inverse problems are solved using LSQR, using an implicit parallel Jacobian operator. Numerical experiments demonstrate the successful recovery of a synthetic 3-D conductivity structure with approximately 700,000 degrees of freedom. The study further discusses computational bottlenecks related to memory consumption and shared-memory scalability arising from the simultaneous storage of multiple sparse matrix factorizations. Possible improvements based on preconditioned iterative solvers and distributed high-performance computing architectures are outlined. The implementation in the Julia programming language is released as open-source software to support reproducible research and further development by the geophysical inversion community.
arXiv:2605.19135v1 Announce Type: new Abstract: Causal representation learning (CRL) seeks to uncover meaningful latent variables and their corresponding causal structure from high-dimensional observational data. Although its significance, CRL identifiability remains a crucial property, as it ensures the recovery of the mechanisms behind the data generation process, and hence the interpretability and robustness of the representation. Proving identifiability in CRL is intrinsically difficult, and we address in this work an even more challenging setting: multimodality. We consider multimodal observed data with a latent partially shared structure. Each modality is generated, through non linear mixing functions, from a specific subset of causal latent variables. Under flexible assumptions and without imposing any parametric distribution on the latent variables, we establish component-wise identifiability guarantees for the causal latent representation. Our identifiability results, furthermore, apply to the undercomplete scenario where we have, for each modality, more observed than latent variables. To instantiate our theoretical analysis, we introduce a Wasserstein-based module to recover the partially shared latent structure. Due to its differentiability, the latter can be easily integrated into all types of architecture, only requiring minimal changes. Extensive experiments on synthetic and realistic datasets validate the superiority of our approach over SOTA methods.
arXiv:2604.18739v3 Announce Type: replace Abstract: Masked diffusion large language models (dLLMs) are a promising alternative to autoregressive generation. While reinforcement learning (RL) methods have recently been adapted to dLLM fine-tuning, their objectives typically depend on sequence-level marginal likelihoods, which are intractable for masked diffusion models. To address this, we derive Discrete Tilt Matching (DTM), a likelihood-free method that recasts dLLM fine-tuning as state-level matching of local unmasking posteriors under reward tilting. DTM takes the form of a weighted cross-entropy objective with explicit minimizer, and admits control variates that improve training stability. On a synthetic maze-planning task, we analyze how DTM's annealing schedule and control variates affect training stability and prevent mode collapse. At scale, fine-tuning LLaDA-8B-Instruct with DTM yields strong gains on Sudoku and Countdown while remaining competitive on MATH500 and GSM8K.
arXiv:2504.09188v2 Announce Type: replace Abstract: This paper introduces the Compliant Explicit Reference Governor (CERG), a modular reference management system that enables robots to interact physically with their environment under provable guarantees. The CERG is an intermediate layer that can be placed between a high-level planner and a low-level controller: it enforces operational constraints and enables smooth transitions between free-motion and contact operations. The CERG ensures safety by limiting the total energy available to the robotic arm at the time of contact. In the absence of contact, however, the CERG does not penalize the system performance. Simulation and hardware experiments validate the CERG on increasingly complex systems.
arXiv:2605.19172v1 Announce Type: new Abstract: Forecasting urban delivery demand becomes substantially more challenging when newly added service regions lack historical records. Existing spatiotemporal forecasters effectively model spatial dependence once sufficient node histories are available. Still, they remain parametric and therefore struggle to recover short-term operational dynamics in cold-start regions. Geospatial embeddings help identify where a region is and what function it serves, yet they do not directly reveal how a similar region behaves under a comparable temporal context. We propose Bridge, a retrieval-augmented spatiotemporal graph framework that combines an inductive contextual graph backbone with a time-aware memory of region-time windows. For each target region, Bridge retrieves future demand patterns from the memory using both regional context and recent dynamics, and refines the backbone forecast through a gated fusion mechanism. To align retrieval with forecasting utility, we further train the retriever with a future-aware objective that favors entries whose future trajectories best match the target. Experiments on four real-world delivery datasets show that Bridge consistently improves over competitive spatiotemporal baselines in both within-city cold-start and cross-city transfer with partial observations. The results show that retrieval augmentation provides a useful operational memory for cold-start urban demand forecasting when parametric graph generalization alone is insufficient.
arXiv:2603.28524v2 Announce Type: replace-cross Abstract: An accurate and efficient numerical electromagnetic model for superconducting qubits is essential for characterizing and minimizing design-dependent dielectric losses. The energy participation ratio (EPR) is the commonly adopted metric used to evaluate these losses, but its calculation presents a severe multiscale computational challenge. Conventional finite element method (FEM) requires 3D volumetric meshing, leading to prohibitive computational costs and memory requirements when attempting to capture singular electric fields at nanometer-thin material interfaces. To address this bottleneck, we propose SesQ, a surface integral equation simulator tailored for the precise simulation of the EPR. By applying discretization on 2D surfaces, deriving a semi-analytical multilayer Green's function, and employing a dedicated non-conformal boundary mesh refinement scheme, SesQ accurately resolves singular edge fields without an explosive growth in the number of unknowns. Validations with analytically solvable models demonstrate that SesQ accelerates capacitance extraction by roughly two orders of magnitude compared to commercial FEM tools. While achieving comparable accuracy for capacitance extraction, SesQ delivers superior precision for EPR calculation. Simulations of practical transmon qubits further reveal that FEM approaches tend to significantly underestimate the EPR. Finally, the high efficiency of SesQ enables rapid iteration in the layout optimization, as demonstrated by minimizing the EPR of the qubit pattern, establishing the simulator as a powerful tool for the automated design of low-loss superconducting quantum circuits.
arXiv:2605.20094v1 Announce Type: new Abstract: Most of the laws of Nature involve derivatives up to the second order. Ostrogradski was the first to seek a formulation of the equations of higher-order derivatives. He extended Hamilton's equations by considering Lagrangians that depend on higher-order derivatives of generalized coordinates. The Hamilton-Ostrogradski formulation served as the basis for later studies with higher-order derivatives. However, the Hamilton-Ostrogradski formalism is rarely discussed in textbooks or the pedagogical literature. This motivated us to show how the Hamilton-Ostrogradski formalism can be applied it to the Pais-Uhlenbeck oscillator. We hope that the approach presented in this work can serve as a basis for discussion in advanced classical mechanics courses.
arXiv:2605.19796v1 Announce Type: new Abstract: This article extends a recently introduced kinetic closure of turbulence by developing its theoretical framework, operational realizations, and validation. In contrast with filtered Navier--Stokes formulations, filtering the Boltzmann equation retains subgrid advective transport under the linear streaming operator, so that unresolved physics is concentrated on the collision side. We show that in the dilute-gas LES and RANS regimes, the main limitation of Boltzmann and BGK-type collision models is not the breakdown of molecular chaos, but the retention of a Markovian collision process at a scale where filtering induces finite temporal correlations in the collision product. In a BGK-type framework, the closure problem is dual: one must infer the filtered fine-grained equilibrium, which is not computable from filtered moments alone, and model the non-Markovian collision dynamics generated by the collision-product covariance. The present framework makes this dual structure explicit and represents the resulting collision-covariance source term through a BGK-like closure built from the subgrid equilibrium residual, with the turbulent relaxation frequency given by a first phenomenological realization. The framework relies on a Chapman--Enskog analysis organized by the reference timescale ratio emerging directly from the nondimensionalization of the kinetic equation and performed in the classical sense, thereby avoiding artificial turbulent scale separations. We show that the Chapman--Enskog structure is not a pure one-parameter Knudsen scaling: the primary ordering is set by the kinetic-to-macroscopic timescale ratio, while higher moments retain an additional Mach dependence through the mixed scaling of particle velocity. The resulting kinetic closures are validated through lattice Boltzmann simulations and compared with the Smagorinsky model and regularization-based collision models.
arXiv:2602.13724v2 Announce Type: replace Abstract: We develop and analyze a model for a flat microbial droplet growing on the surface of a three-dimensional viscous fluid. The model describes growth-induced stresses at the fluid surface, density variations in the bulk due to nutrient consumption, and the resulting fluid flows that arise. We reformulate this free-boundary problem as a system of integro-differential equations defined solely on the microbial domain. From this formulation, we identify an axisymmetric solution corresponding to a radially expanding disk and analyze its morphological stability. We find that growth forces stabilize the axisymmetric solution while buoyancy forces destabilize it. We connect these findings to experimental observations.
arXiv:2506.18852v2 Announce Type: replace Abstract: Mechanistic interpretability (MI) aims to explain how neural networks work by uncovering their underlying mechanisms. As the field grows in influence, it is increasingly important to examine not just models themselves, but the assumptions, concepts and explanatory strategies implicit in MI research. We argue that mechanistic interpretability needs philosophy as an ongoing partner in clarifying its concepts, refining its methods, and navigating the epistemic and ethical complexities of interpreting AI systems. There is significant unrealised potential for progress in MI to be gained through deeper engagement with philosophers and philosophical frameworks. Taking three open problems from the MI literature as examples, this paper illustrates the value philosophy can add to MI research, and outlines a path toward deeper interdisciplinary dialogue.
arXiv:2605.19133v1 Announce Type: new Abstract: Self-supervised learning (SSL) is now a standard way to pretrain medical image models, but performance is still mostly judged by downstream accuracy. For safety-critical screening tasks such as diabetic retinopathy grading, this is not enough: a model must also know when its predictions are unreliable and defer uncertain cases for clinical review. In this work, we examine how the length of SSL pretraining influences calibrated confidence and confidence-based abstention. We evaluate multiple SSL checkpoints under a fixed fine-tuning protocol and assess calibrated confidence, coverage, selective accuracy, and selective macro-F1. Across datasets and data regimes, SSL pretraining improves selective prediction compared to training from scratch. Unlike prior SSL studies that primarily evaluate downstream accuracy or AUROC, we analyze how SSL pretraining duration influences confidence behavior under calibrated confidence-based abstention. However, once accuracy saturates, selective performance can still change markedly across checkpoints, and longer pretraining does not consistently improve reliability. These results underscore the importance of abstention-aware evaluation and suggest that pretraining length should be treated as an important reliability-related design choice rather than only a computational detail. Code is available at GitHub.
arXiv:2605.19259v1 Announce Type: new Abstract: We propose ERFSL, an efficient reward function searcher using large language models (LLMs) for custom-environment, multi-objective learning-based methods (LB). ERFSL generates reward components based on explicit user requirements, rectifies them using a reward critic, and iteratively optimizes the weights of these components based on textual context generated by the training log analyzer. Applied to a simulation-based benchmark task, the reward critic corrects reward codes with only one feedback iteration per requirement, and the reward weight initializer acquires diverse reward functions within the Pareto set. Even when a weight is off by a factor of 500, an average of only 5.2 iterations is needed to meet user requirements. The approach works adequately with GPT-4o mini and does not require advanced understanding capabilities.
arXiv:2605.19123v1 Announce Type: new Abstract: Cryptographic primitives such as stream ciphers,Pseudorandom Number Generators (PRNGs), and block cipher modes produce sequences that are designed to be statistically indistinguishable from random data. As a result, the traditional evaluation techniques therefore rely primarily on statistical randomness tests to assess the quality of generated sequences. While these tests verify global statistical properties, they do not address whether structural characteristics of sequences can reveal information about the underlying generator. In this paper, we introduce a stringology-based fingerprinting, (SBF) framework for the structural analysis of cryptographic sequences. The proposed SBF framework interprets cryptographic outputs as symbolic strings and applies pattern-based feature extraction to capture structural statistics such as substring frequency distributions, recurrence patterns, and entropy characteristics. These structural features are aggregated into fingerprint vectors that characterize sequence generators. The experimental evaluation is conducted using datasets composed of Cipher-Generated Sequences (CGS) and Uniformly Random Sequences (URS). The results demonstrate that stringology-based pattern analysis can reveal measurable structural signatures across different sequence sources. Although these signals do not imply practical cryptographic weaknesses, they provide an additional analytical perspective for evaluating the structural behavior of cryptographic generators.
arXiv:2510.00600v2 Announce Type: replace Abstract: Using Large Language Models to produce intermediate thoughts, a.k.a. Chain-of-thought (CoT), before providing an answer has been a successful recipe for solving complex language tasks. In robotics, similar embodied CoT strategies, generating thoughts before actions, have also been shown to lead to improved performance when using Vision-Language-Action models (VLAs). As these techniques increase the length of the model's generated outputs to include the thoughts, the inference time is negatively affected. Delaying an agent's actions in real-world executions, as in robotic manipulation settings, strongly affects the usability of a method, as tasks require long sequences of actions. However, is the generation of long chains-of-thought a strong prerequisite for achieving performance improvements? In this work, we explore the idea of Hybrid Training (HyT), a framework that enables VLAs to learn from thoughts and benefit from the associated performance gains, while enabling the possibility to leave out CoT generation during inference. Furthermore, by learning to conditionally predict a diverse set of outputs, HyT supports flexibility at inference time, enabling the model to either predict actions directly, generate thoughts or follow instructions. We evaluate the proposed method in a series of simulated benchmarks and real-world experiments.
arXiv:2605.19119v1 Announce Type: new Abstract: Existing neural combinatorial optimization solvers frame solution search as imitation of optimal decisions, inherently limiting their utility to single-objective minimization and static constraints. We propose GOAL, a conditioned diffusion solver over relational graph representations that enables controllable decision generations by conditioning on human-specified objectives. We introduce a heterogeneous graph encoding in which distinct edge types, corresponding to different classes of constraints, define the message passing structure of the graph neural network, which allows information to propagate selectively according to the ontology of each constraint. GOAL is instantiated and evaluated on three canonical scheduling benchmarks of various constraint complexity: the Flow Shop Problem (FSP), the Job Shop Scheduling Problem (JSP), and the Flexible Job Shop Scheduling Problem (FJSP). Generalization is demonstrated across structurally distinct constraint regimes and problem types without architectural modification. On all three benchmarks, GOAL achieves 100% solution feasibility and near-zero MAPE (below 0.20%) on multiple objectives for problem sizes up to 20 jobs and 60 operations, outperforming NSGA-II and MOEA/D in both solution quality and inference speed by up to 25x.
arXiv:2605.19102v1 Announce Type: new Abstract: Large Language Models (LLMs) can generate code from natural language, but their performance is highly sensitive to prompt formulation. We propose a reinforcement-learning-based framework that models prompt refinement as a sequential decision-making problem. A Proximal Policy Optimization (PPO) agent iteratively improves prompts using a hybrid action space that combines direct generation, genetic lexical mutation and semantic rewriting, guided by shaped rewards derived from unit-test feedback. We evaluate the framework on MBPP+, HumanEval+, and APPS using CodeT5+, CodeLLaMA, and DeepSeek-Coder as frozen code generators. On the 500-task MBPP+ test set, the PPO agent achieves strict Pass@1 scores of 57.58%, 64.80%, and 85.50%, respectively, outperforming EPiC, Reflexion, and Random-Hybrid. Soft-Pass@1 reaches 67.90%, 73.10%, and 88.20%, respectively. Similar improvements are observed on HumanEval+ and APPS across all backbone models. The results demonstrate that reinforcement learning with shaped test-driven rewards improves functional correctness in LLM-based code generation.
arXiv:2605.19792v1 Announce Type: new Abstract: Visually-grounded language models (VLMs) are highly effective in linking visual and textual information, yet they often struggle with basic classification and localization tasks. While classification mechanisms have been studied more extensively, the processes that support object localization remain poorly understood. In this work, we investigate two representative families, LLaVA-1.5 and InternVL-3.5, using a suite of mechanistic interpretability tools, including token ablations, attention knockout, and causal mediation analysis. We find that localization is driven by a containerization mechanism in which object-aligned tokens define the spatial extent of the object, while the semantic arrangement of tokens within those boundaries is largely irrelevant to the predicted box. Only a very small set of attention heads mediates the causal effect for both classification and localization, concentrating in early-mid layers for LLaVA and mid-late layers for InternVL. The two tasks share some early processing but ultimately depend on largely distinct specialized heads. Overall, we provide the first layer- and head-level account of localization in VLMs, revealing narrow computational pathways that can guide future model design and grounding objectives.
arXiv:2605.20024v1 Announce Type: new Abstract: This study examines how parents of LGBTQ+ individuals in South Korea navigate the emotional rupture fueled by fear, isolation, and disorientation after learning their children's queer identity, encounter queer-related (mis)information as a way of coping with this emotional toll, and come to listen to queer realities relationally. Through this process, we highlight how parents reconstruct their identities as supportive parents, which reshapes their informating practices, making them more critical in assessing queer-related (mis)information, developing strategies to protect themselves from harmful narratives, and actively challenging misinformation to support others navigating similar experiences. This work contributes to CSCW by (1) foregrounding parents of LGBTQ+ individuals, an underrepresented yet critical stakeholder group in Queer HCI; (2) demonstrating how identity reconfiguration following a trauma-healing process could transform information practices; and (3) arguing that addressing misinformation requires attention beyond individual fact-based discerning to account for its relational, cultural, and emotional dimensions. Further, we invite CSCW scholars to reconsider the balance between abstracting and humanizing information, explore future design possibilities for parents of LGBTQ+ children, and reflect on the role of researchers as participants in collective research communities fueled by care.
arXiv:2605.20035v1 Announce Type: new Abstract: Omni-modal large language models (om-LLMs) achieve unified audio-visual understanding by encoding video and audio into temporally aligned token sequences interleaved at the window level. However, processing these dense non-textual tokens throughout the LLM incurs substantial computational overhead. Although training-free token selection can reduce this cost, existing methods either focus on visual-only inputs or prune om-LLM tokens only before the LLM with fixed per-modality ratios, failing to capture how cross-modal token importance evolves across layers. To address this limitation, we first analyze the layer-wise token dependency of om-LLMs. We find that visual and audio dependencies follow a block-wise pattern and gradually weaken with depth, indicating that many late-layer non-textual tokens become redundant after cross-modal fusion. Motivated by this observation, we propose SEATS, a training-free, stage-adaptive token selection method for efficient om-LLM inference. Before the LLM, SEATS removes spatiotemporal redundancy via attention-weighted diversity selection. Inside the LLM, it progressively prunes tokens across blocks and dynamically allocates the retention budget from temporal windows to modalities using query relevance scores. In late layers, it removes all remaining non-textual tokens once cross-modal fusion is complete. Experiments on Qwen2.5-Omni and Qwen3-Omni demonstrate that SEATS effectively improves inference efficiency. Retaining only 10% of visual and audio tokens, it achieves a 9.3x FLOPs reduction and a 4.8x prefill speedup while preserving 96.3% of the original performance.
arXiv:2407.13193v4 Announce Type: replace Abstract: Large language models (LLMs) have achieved strong empirical performance in various fields, benefiting from their huge amount of parameters that store knowledge. However, LLMs still suffer from several key issues, such as hallucination problems, knowledge update issues, and lacking domain-specific expertise. The appearance of retrieval-augmented generation (RAG), which leverages an external knowledge base to augment LLMs, mitigates these limitations. This paper presents a systematic review of RAG techniques for natural language processing (NLP), with a focus on retrievers and retrieval fusions. We introduce a novel taxonomy of retrieval fusions, such as query-based, logits-based, latent, and parametric fusion, and provide structured comparisons across accessibility, efficiency, and use cases. The paper further examines RAG applications across diverse NLP tasks, discusses evaluation methodologies and benchmark limitations, and analyzes training paradigms with and without knowledge base updates. Finally, we explore industrial deployment considerations and identify emerging challenges and future directions, including security, efficiency, and graph-based retrieval.
arXiv:2605.19762v1 Announce Type: new Abstract: Code has become a standard component of modern foundation language model (LM) training, yet its role beyond programming remains unclear. We revisit the claim that code improves reasoning through controlled pretraining experiments on a 10T-token corpus with fine-grained domain separation. Our findings are threefold. First, when code is restricted to standalone executable programs and Code-NL data are controlled for, code substantially improves programming ability but does not act as a general reasoning enhancer; instead, it competes with knowledge-intensive tasks, especially complex mathematical reasoning. Second, the reasoning gains often attributed to code are better explained by cross-domain structured reasoning traces, such as code-text and math-text mixtures, rather than by executable code alone. Third, increasing the density of structured math-domain samples within a fixed math budget yields substantial gains on difficult mathematical reasoning while largely preserving programming performance, suggesting that cognitive scaffolds offer a targeted way to mitigate cross-domain trade-offs. Finally, routing analyses show that data-composition effects are reflected in expert-activation patterns, providing mechanism-level evidence for competitive and synergistic interactions across domains. Our results clarify which data characteristics transfer across capability dimensions and point to more precise data-centric optimization strategies.
arXiv:2404.07106v2 Announce Type: replace Abstract: Point cloud completion aims to generate a complete and high-fidelity point cloud from an initially incomplete and low-quality input. A prevalent strategy involves leveraging Transformer-based models to encode global features and facilitate the reconstruction process. However, the adoption of pooling operations to obtain global feature representations often results in the loss of local details within the point cloud. Moreover, the attention mechanism inherent in Transformers introduces additional computational complexity, rendering it challenging to handle long sequences effectively. To address these issues, we propose 3DMambaComplete, a point cloud completion network built on the novel Mamba framework. It comprises three modules: HyperPoint Generation encodes point cloud features using Mamba's selection mechanism and predicts a set of Hyperpoints. A specific offset is estimated, and the down-sampled points become HyperPoints. The HyperPoint Spread module disperses these HyperPoints across different spatial locations to avoid concentration. Finally, a deformation method transforms the 2D mesh representation of HyperPoints into a fine-grained 3D structure for point cloud reconstruction. Extensive experiments conducted on various established benchmarks demonstrate that 3DMambaComplete surpasses state-of-the-art point cloud completion methods, as confirmed by qualitative and quantitative analyses.