arXiv:2407.17277v3 Announce Type: replace
Abstract: We introduce data to predictive control, D2PC, a framework to facilitate the design of robust and predictive controllers from data. The proposed framework is designed for discrete-time stochastic linear systems with output measurements and provides a principled design of a predictive controller based on data. The framework builds on a parameter identification method based on the Expectation-Maximization algorithm, which incorporates pre-defined structural constraints. An asymptotic approximation is leveraged to quantify the uncertainty in the parameter estimates. As the main contributions, a robust control and predictive control design are proposed tailored to the uncertainty characterization resulting from the identification. In particular, a strategy to synthesize robust dynamic output-feedback controllers is presented. Furthermore, a predictive control scheme that guarantees recursive feasibility and satisfaction of chance constraints is developed. This framework marks a significant advancement in integrating data-driven models into robust and predictive control designs. We demonstrate the efficacy of D2PC through a numerical example involving a $10$-dimensional spring-mass-damper system.
Science Journals
arXiv:2407.20595v5 Announce Type: replace
Abstract: Authorship attribution asks whether two pieces of text share a writer, but topical confound makes the task deceptively easy: two authors covering the same topic may look more alike than one author covering two topics. Scholarly prose offers a natural remedy, academic writers produce multiple papers on related but distinct topics while maintaining consistent stylistic habits. We introduce HALvest, a 17-billion-token multilingual corpus of open-access academic papers, and its English contrastive derivative HALvest-Contrastive, where same-author passages are drawn from distinct papers within a disciplinary field to minimize topical overlap. We validate our benchmark by showing that a strong lexical baseline collapses once topical shortcuts are removed. On this same benchmark, we revisit how authorship is scored. Standard systems compress each document into a single vector. We instead keep a sequence of vectors and compare them with late interaction, then propose patch-level late interaction, which groups neighboring tokens into patches before matching. Matching at the sequence level greatly improves performance over the single-vector baseline, but the optimal interaction granularity is subtle.
arXiv:2409.00346v5 Announce Type: replace
Abstract: In medical image segmentation, specialized computer vision techniques, notably transformers grounded in attention mechanisms and residual networks employing skip connections, have been instrumental in advancing performance. Nonetheless, previous models often falter when segmenting small, irregularly shaped tumors. To this end, we introduce SMAFormer, an efficient, Transformer-based architecture that fuses multiple attention mechanisms for enhanced segmentation of small tumors and organs. SMAFormer can capture both local and global features for medical image segmentation. The architecture comprises two pivotal components. First, a Synergistic Multi-Attention (SMA) Transformer block is proposed, which has the benefits of Pixel Attention, Channel Attention, and Spatial Attention for feature enrichment. Second, addressing the challenge of information loss incurred during attention mechanism transitions and feature fusion, we design a Feature Fusion Modulator. This module bolsters the integration between the channel and spatial attention by mitigating reshaping-induced information attrition. To evaluate our method, we conduct extensive experiments on various medical image segmentation tasks, including multi-organ, liver tumor, and bladder tumor segmentation, achieving state-of-the-art results. Code and models are available at: https://github.com/lzeeorno/SMAFormer.
arXiv:2409.02416v2 Announce Type: replace
Abstract: Motivated by the Bures distance, we introduce a new family of distances, \emph{relative translation invariant Wasserstein distances}, denoted by $RW_p$, as an extension of the classical Wasserstein distances $W_p$ for $p \in [1, +\infty)$. We establish that $RW_p$ defines a valid metric and demonstrate that this type of metric is more intrinsic than the classical Wasserstein distance. A bi-level algorithm is designed to compute the general $RW_p$ distance between arbitrary discrete distributions. Moreover, when $p = 2$, we show that the optimal coupling matrix is invariant under distributional translation in the discrete setting, and we further propose two algorithms, the $\mathrm{RW}_2$-LP algorithm and the $\mathrm{RW}_2$-Sinkhorn algorithm, to improve the numerical stability of computing $W_2$ distance and the optimal coupling matrix solutions. Finally, we conduct three experiments to validate our theoretical results and algorithms. The first two experiments report that the $\mathrm{RW}_2$-LP algorithm and the $\mathrm{RW}_2$-Sinkhorn algorithm, both with and without normalization, can significantly reduce the numerical errors compared to standard algorithms. The third experiment shows that $RW_p$ algorithms are computationally scalable and applicable to the retrieval of similar thunderstorm patterns in practical applications.
arXiv:2409.08379v4 Announce Type: replace
Abstract: Large Language Models (LLMs) are reshaping knowledge work, yet their impact on voluntary, self-guided open innovation forums (contributors choose tasks without managerial direction) may differ fundamentally from effects observed in organizational settings. We study this question in open-source software development, where individuals' contributions collectively drive innovation at a community level. Unlike product innovation, where typologies for classifying innovation are well established, knowledge work in open-source settings calls for a distinction grounded in the cognitive demand a task places on the contributor. Burgeoning literature distinguishes substantive contributions, which require creative problem formulation to introduce new functionality, from incremental contributions, which draw on comprehension of existing code to maintain and refine it. We exploit a natural experiment around GitHub Copilot's launch in October 2021, where Copilot supported languages like Python while not supporting R for business reasons, creating an exogenous partition between otherwise comparable ecosystems. Using three complementary identification strategies and two classification approaches, we find that Copilot availability increases open-source contributions by 28 to 40 percent. The increase in incremental contributions is significantly larger than the increase in substantive contributions across all specifications. This disparity is more pronounced in projects with higher activity levels and widens following a model upgrade: LLMs function more effectively when existing context helps define the problem and constrain solutions, tilting collaborative innovation toward exploitation of established codebases rather than exploration of new functionality. This paper provides a rare instance of causal field evidence on LLM effects, given the speed at which GenAI has exploded across the knowledge economy.
arXiv:2605.24544v1 Announce Type: new
Abstract: We present JAX-SCM v1.0, an open-source atmospheric single-column model for boundary layer research, implemented in Python using the JAX computing library. The model solves for horizontal wind, potential temperature, and specific humidity, combined with prognostic turbulent kinetic energy and turbulent statistics parameterized by the Mellor-Yamada-Nakanishi-Niino level-2.5 (MYNN-2.5) turbulence closure. We verify the implementation against three well-established benchmark cases covering neutral (turbulent Ekman layer), stable (GABLS1), and convective (Wangara Day 33) conditions. Close agreement with reference solutions is demonstrated across all regimes. By building on JAX, the model benefits from just-in-time compilation and native GPU support. While JAX-SCM is not yet fully differentiable, basing it on JAX also lays the foundation for future integration with machine learning components. The model is designed for simplicity and modularity, lowering the barrier to entry for users and developers alike.
arXiv:2409.09953v3 Announce Type: replace
Abstract: Out-of-distribution (OOD) detection targets to detect and reject test samples with semantic shifts, to prevent models trained on in-distribution (ID) dataset from producing unreliable predictions. Existing works only extract the appearance features on image datasets, and cannot handle dynamic multimedia scenarios with much motion information. Therefore, we target a more realistic and challenging OOD detection task: OOD action detection (ODAD). Given an untrimmed video, ODAD first classifies the ID actions and recognizes the OOD actions, and then localizes ID and OOD actions. To this end, in this paper, we propose a novel Uncertainty-Guided Appearance-Motion Association Network (UAAN), which explores both appearance features and motion contexts to reason spatial-temporal inter-object interaction for ODAD.Firstly, we design separate appearance and motion branches to extract corresponding appearance-oriented and motion-aspect object representations. In each branch, we construct a spatial-temporal graph to reason appearance-guided and motion-driven inter-object interaction. Then, we design an appearance-motion attention module to fuse the appearance and motion features for final action detection. Experimental results on two challenging datasets show that UAAN beats state-of-the-art methods by a significant margin, illustrating its effectiveness.
arXiv:2409.10184v2 Announce Type: replace
Abstract: We present an efficient moment-based perturbation scheme for evaluating polarizability tensors of small molecules at a fraction of the computational cost of conventional energy-based approaches. Rather than applying explicit electric fields, the method perturbs the molecular charge density using strategically arranged external point charges, as in QM/MM simulations, and extracts polarizability tensors from finite differences of multipole moments extrapolated to the zero-perturbation limit. C-Pol is implemented as a backend-agnostic Python package requiring only that the host quantum chemistry code support external point-charge potentials. We demonstrate interfaces with three codes spanning complementary numerical approaches: GPAW (real-space grids), NWChem, and PySCF (atom-centered basis sets with complete basis set extrapolation). Validation against Gaussian 16 energy-based reference calculations across molecules representing all 19 commonly occurring point groups yields agreement within 3%, with the largest deviations confined to tensor components that are numerically small and contribute negligibly to the electrostatic potential. Finite-difference polarizabilities computed via C-Pol are further shown to agree with analytic coupled-perturbed Hartree-Fock values to within 0.02 a.u. across all tested basis sets, at comparable computational cost for medium-sized systems and with more favorable scaling for larger ones. The package provides a practical and transferable route to high-quality multipole polarizability data for force field development, polarizable embedding, and machine-learning training sets.
arXiv:2409.19727v3 Announce Type: replace
Abstract: Deep Neural Networks (DNNs) are often over-parameterized for their tasks and can be compressed quite drastically by removing weights, a process called pruning. We investigate the impact of different pruning techniques on the classification performance and interpretability of GoogLeNet. We systematically apply unstructured and structured pruning, as well as connection sparsity (pruning of input weights) methods to the network and analyze the outcomes regarding the network's performance on the validation set of ImageNet. We also compare different retraining strategies, such as iterative pruning and one-shot pruning. We find that with sufficient retraining epochs, the performance of the networks can approximate the performance of the default GoogLeNet - and even surpass it in some cases. To assess interpretability, we employ the Mechanistic Interpretability Score (MIS) developed by Zimmermann et al. . Our experiments reveal that there is no significant relationship between interpretability and pruning rate when using MIS as a measure. Additionally, we observe that networks with extremely low accuracy can still achieve high MIS scores, suggesting that the MIS may not always align with intuitive notions of interpretability, such as understanding the basis of correct decisions.
arXiv:2503.19605v5 Announce Type: replace
Abstract: Understanding and certifying the generalization performance of machine learning algorithms -- i.e. obtaining theoretical estimates of the test error from the training error -- is a central theme of statistical learning theory. Among the many complexity measures used to derive such guarantees, Rademacher complexity yields sharp, data-dependent bounds that apply well beyond classical VC-dimension theory. In this study, we formalize the generalization error bound by Rademacher complexity in Lean 4, building on measure-theoretic probability theory available in the Mathlib library. Our development provides a mechanically-checked pipeline from the definitions of empirical and expected Rademacher complexity, through a formal symmetrization argument and a bounded-differences analysis, to high-probability uniform deviation bounds via a formally proved McDiarmid inequality. A key technical contribution is a reusable mechanism for lifting results from countable hypothesis classes (where measurability of suprema is straightforward in Mathlib) to separable topological index sets via a reduction to a countable dense subset. As worked applications of the abstract theorem, we mechanize standard empirical Rademacher bounds for linear predictors under $\ell_2$ and $\ell_1$ regularizations, and we also formalize a Dudley-type entropy integral bound based on covering numbers and a chaining construction.
arXiv:2503.23670v2 Announce Type: replace
Abstract: Inferring signed distance functions (SDFs) from sparse point clouds remains a challenge in surface reconstruction. The key lies in the lack of detailed geometric information in sparse point clouds, which is essential for learning a continuous field. To resolve this issue, we present a novel approach that learns a dynamic deformation network to predict SDFs in an end-to-end manner. To parameterize a continuous surface from sparse points, we propose a bijective surface parameterization (BSP) that learns the global shape from local patches. Specifically, we construct a bijective mapping for sparse points from the parametric domain to 3D local patches, integrating patches into the global surface. Meanwhile, we introduce grid deformation optimization (GDO) into the surface approximation to optimize the deformation of grid points and further refine the parametric surfaces. Experimental results on synthetic and real scanned datasets demonstrate that our method significantly outperforms the current state-of-the-art methods. Project page: https://takeshie.github.io/Bijective-SDF
arXiv:2504.04197v2 Announce Type: replace
Abstract: Smoothed analysis is a method for analyzing the performance of algorithms, used especially for those algorithms whose running time in practice is significantly better than what can be proven through worst-case analysis. Spielman and Teng (STOC '01) introduced the smoothed analysis framework of algorithm analysis and applied it to the simplex method. Given an arbitrary linear program with $d$ variables and $n$ inequality constraints, Spielman and Teng proved that the simplex method runs in time $O(\sigma^{-30} d^{55} n^{86})$, where $\sigma > 0$ is the standard deviation of Gaussian distributed noise added to the original LP data. Spielman and Teng's result was simplified and strengthened over a series of works, with the current strongest upper bound being $O(\sigma^{-3/2} d^{13/4} \log(n)^{7/4})$ pivot steps due to Huiberts, Lee and Zhang (STOC '23). We prove that there exists a simplex method whose smoothed complexity is upper bounded by $O(\sigma^{-1/2} d^{11/4} \log(n)^{7/4})$ pivot steps. Furthermore, we prove a matching high-probability lower bound of $\Omega( \sigma^{-1/2} d^{1/2}\ln(4/\sigma)^{-1/4})$ on the combinatorial diameter of the feasible polyhedron after smoothing, on instances using $n = \lfloor (4/\sigma)^d \rfloor$ inequality constraints. This lower bound indicates that our algorithm has optimal noise dependence among all simplex methods, up to polylogarithmic factors.
arXiv:2504.04639v4 Announce Type: replace
Abstract: It is an open question whether the search and decision versions of promise CSPs are equivalent. Most known algorithms for PCSPs solve only their \emph{decision} variant, and it is unknown whether they can be adapted to solve \emph{search} as well. The main approaches, called BLP, AIP and BLP+AIP, handle a PCSP by finding a solution to a relaxation of some integer program. We prove that rounding those solutions to a proper search certificate can be as hard as any problem in the class TFNP. In other words, these algorithms are ineffective for search. Building on the algebraic approach to PCSPs, we find sufficient conditions that imply ineffectiveness for search. Our tools are tailored to algorithms that are characterized by minions in a suitable way, and can also be used to prove undecidability results for meta-problems. This way, we show that the families of templates solvable via BLP, AIP, and BLP+AIP are undecidable.
Using the same techniques we also analyze several algebraic conditions that are known to guarantee the tractability of finite-template CSPs. We prove that several meta-problems related to cyclic polymorphims and WNUs are undecidable for PCSPs. In particular, there is no algorithm deciding whether a finite PCSP template (1) admits cyclic a polymorphism, (2) admits a WNU.
arXiv:2504.11961v3 Announce Type: replace
Abstract: Zero-knowledge (ZK) circuits enable privacy-preserving computations and are central to many cryptographic protocols. Systems like Circom simplify ZK development by combining witness computation and circuit constraints in one program. However, even small errors can compromise security of ZK programs -- under-constrained circuits may accept invalid witnesses, while over-constrained ones may reject valid ones. Static analyzers are often imprecise with high false positives, and formal tools struggle with real-world circuit scale. Additionally, existing tools overlook several critical behaviors, such as intermediate computations and program aborts, and thus miss many vulnerabilities. Our theoretical contribution is the Trace-Constraint Consistency Test (TCCT), a foundational, language-independent formulation of ZK circuit bugs. TCCT provides a unified semantics that subsumes prior definitions and captures both under- and over-constrained vulnerabilities, exposing the full space of ZK bugs that elude prior tools. Our systems contribution is zkFuzz, a novel program mutation-based fuzzing framework for detecting TCCT violations. zkFuzz systematically mutates the computational logic of Zk programs guided by a novel fitness function, and injects carefully crafted inputs using tailored heuristics to expose bugs. We evaluated zkFuzz on 452 real-world ZK circuits written in Circom, a leading programming system for ZK development. zkFuzz successfully identified 85 bugs, including 59 zero-days-39 of which were confirmed by developers and \nfixed fixed, including bugs undetectable by prior works due to their fundamentally limited formulations, earning thousands of bug bounties. Our preliminary research on Noir, another emerging DSL for ZK circuit, also demonstrates the feasibility of zkFuzz to support multiple DSLs.
arXiv:2505.02129v3 Announce Type: replace
Abstract: Organizing large-scale resources in a multidimensional semantic space is an approach to efficiently managing and querying resources from different semantic dimensions. To support advanced applications, this paper proposes a resource space model for aggregation query on subspaces defined by a range within the partial order on the coordinate trees representing each dimension, where each point in the subspace contains resources aggregated along the paths of the partial order relations on the coordinate trees and the aggregated resources at each point can be measured, ranked and selected by applications. To efficiently locate non-empty points in a large subspace, an approach to generating graph index is proposed to build partial order relations on coordinates of dimensions to enable a subspace query to reach non-empty points through indexing links and aggregate resources along indexing paths to their super points. Generating such an index is costly as the number of children of an indexing node can be large so that the total number of indexing nodes can be very large (exponentially growing with the number of dimensions and scale of dimensions). The proposed approach adopts the a set of strategies to reduce the cost. Analysis and experiments show the effectiveness of the generated index in supporting subspace aggregation query.
arXiv:2505.02761v4 Announce Type: replace
Abstract: Reliable broadcast (RBC) is a key primitive in fault-tolerant distributed systems, and improving its efficiency can benefit a wide range of applications. This work focuses on signature-free RBC protocols, which are particularly attractive due to their computational efficiency. Existing protocols in this setting incur an optimal 3 steps to reach a decision while tolerating up to $f < n/3$ Byzantine faults, where $n$ is the number of parties. In this work, we propose an optimistic RBC protocol that maintains the $f < n/3$ fault tolerance but achieves termination in just 2 steps under certain optimistic conditions--when at least $\lceil \frac{n+2f-2}{2} \rceil$ non-broadcaster parties behave honestly. We also prove a matching lower bound on the number of honest parties required for 2-step termination.
We show that our latency-reduction technique generalizes beyond RBC and applies to other primitives such as asynchronous verifiable secret sharing (AVSS) and asynchronous verifiable information dispersal (AVID), enabling them to complete in 2 steps under similar optimistic conditions.
To highlight the practical impact of our RBC protocol, we integrate it into Sailfish++, a new signature-free, post-quantum secure DAG-based Byzantine fault-tolerant (BFT) consensus protocol. Under optimistic conditions, this protocol achieves a commit latency of 3 steps--matching the performance of the best signature-based protocols. Our experimental evaluation shows that our protocol significantly outperforms existing post-quantum secure and signature-based protocols, even on machines with limited CPU resources. In contrast, signature-based protocols require high CPU capacity to achieve comparable performance. We have open-sourced our Rust implementation of Sailfish++ to facilitate reproducible results.
arXiv:2505.05304v3 Announce Type: replace
Abstract: Small-scale intermittency is studied as the deviation of the probability distributions of pseudodissipation, dissipation and enstrophy in turbulence from those of a Gaussian random velocity field. This deviation is quantified using Kullback-Leibler (KL) divergence between the two distributions, directly measuring turbulence-induced intermittency separated from purely kinematic effects. Using direct numerical simulation data of forced isotropic turbulence over a wide range of Taylor Reynolds numbers ($Re_{\lambda}$), we characterize the $Re_{\lambda}$ dependence of small-scale intermittency via KL divergence and uncertainty via Shannon entropy, identifying distinct behavioral regimes. Small-scale uncertainty exhibits a non-monotonic dependence on $Re_{\lambda}$: despite continuously growing variability, entropy decays above a certain Reynolds number, suggesting a fundamental change in the statistical nature of the small scales. Turbulence-induced intermittency grows logarithmically with Reynolds number in contrast to the commonly reported power-law scaling, implying that turbulence shows a diminishing growth rate of intermittency at higher Reynolds numbers. Finally, we uncover an emergent symmetry: turbulence dynamics is shown to generate nearly equal intermittency in dissipation rate and enstrophy, challenging the prevailing assumption of asymmetry between strain-rate and vorticity dynamics.
arXiv:2505.05437v4 Announce Type: replace
Abstract: A fundamental question in biology is to understand how patterns and shapes emerge from the collective interplay of large numbers of cells. Cells forming two-dimensional epithelial tissues behave as active materials that undergo remodeling and spontaneous shape changes. Focusing on the fly wing as a model system, we find that the cellular packing in the wing epithelium transitions from a disordered packing to an ordered, crystalline packing. While previous studies propose a role of tissue shear flow in establishing the ordered cell packing in the fly wing, we reveal a role of cell size heterogeneity. Indeed, we find that even if tissue shear have been inhibited, cell packings in the fruit fly wing epithelium transition from disordered to an ordered packing. We propose that the transition is controlled by the cell size heterogeneity, which is quantified by the cell size polydispersity. To explore the role of cell size polydispersity in controlling cellular packings, we implement polydispersity in a vertex model of epithelial tissues. Through numerical simulations of this model, we show that there is a critical value of cell size polydispersity above which cellular packings are disordered and below which they form a crystalline packing. By analyzing experimental data, we find that cell size polydispersity decreases during fly wing development. The observed dynamics of tissue crystallisation is consistent with the slow ordering kinetics we observe in the vertex model. Therefore, although tissue shear does not control the transition, it significantly enhances the rate of tissue-scale ordering by facilitating alignment of locally ordered crystallites. Our results identify cell size heterogeneity as a control parameter, in both the vertex model and the fruit fly wing epithelium, controlling the transition between ordered and disordered cellular packings.
arXiv:2505.08155v4 Announce Type: replace
Abstract: Complex Query Answering (CQA) is a crucial reasoning task over Knowledge Graphs (KGs), which aims to answer first-order logical queries from incomplete KGs. While existing neural-symbolic methods achieve strong performance, they face significant complexity bottlenecks: quadratic data complexity scaling with the number of entities, and NP-hard query complexity for cyclic queries. Consequently, these approaches struggle to scale effectively to large knowledge graphs and complex queries. To address these limitations, we propose an efficient and scalable symbolic search method comprising two key components: (1) constraint strategies that drastically reduce the variable search domain, lowering data complexity; and (2) a local search algorithm that approximately solves NP-hard cyclic queries. Experiments on various CQA benchmarks demonstrate that, for tree-form queries, our method achieves 97% relative MRR with a 10$\times$ speedup using only 10% of the search space. Furthermore, it demonstrates robust performance on complex cyclic queries and large-scale KGs, effectively alleviating efficiency and scalability challenges. Our code is provided in https://github.com/HKUST-KnowComp/NLISA_KDD2026.
arXiv:2505.11758v2 Announce Type: replace
Abstract: Few-shot adaptation of vision-language models remains fundamentally limited by how negative class signals are handled at inference. Existing methods apply uniform negative suppression across all queries, ignoring that the most damaging confusions are query-specific and shift with support-set geometry. We introduce SCAN (Selective Confusion-Aware Negatives), a framework that addresses this gap through three targeted contributions. In inference, query-adaptive negative routing restricts suppression to the top-K most confusable classes per query, requiring zero additional parameters. Generic negative text templates are replaced with LLM-bootstrapped contrastive prompts that describe discriminative attributes between confusable class pairs, sharpening the textual decision boundary where it matters most. A parameter-free adaptive fusion weight estimated from support-set Fisher discriminability removes the need for manual tuning of the vision-language trade-off. Evaluated across 11 standard benchmarks, SCAN consistently outperforms prior prompt-based and adapter-based methods by an average of 4.61% at 16-shot, with gains of up to 7.70% on fine-grained datasets where inter-class confusion is most severe. SCAN also generalizes strongly under distribution shift, improving by 2.95% on average across four ImageNet OOD variants, and maintains robust performance under significant label noise, with accuracy under 50% label corruption still exceeding the clean baseline of the strongest competing method.
arXiv:2505.11788v2 Announce Type: replace
Abstract: To support emerging language-based applications using dispersed and heterogeneous computing resources, the hybrid language model (HLM) offers a promising architecture, where an on-device small language model (SLM) generates draft tokens that are validated and corrected by a remote large language model (LLM). However, the original HLM suffers from substantial communication overhead, as the LLM requires the SLM to upload the full vocabulary distribution for each token. Moreover, both communication and computation resources are wasted when the LLM validates tokens that are highly likely to be accepted. To overcome these limitations, we propose communication-efficient and uncertainty-aware HLM (CU-HLM). In CU-HLM, the SLM transmits truncated vocabulary distributions only when its output uncertainty is high. We validate the feasibility of this opportunistic transmission by discovering a strong correlation between SLM's uncertainty and LLM's rejection probability. Furthermore, we theoretically derive optimal uncertainty thresholds and optimal vocabulary truncation strategies. Simulation results show that, compared to standard HLM, CU-HLM achieves up to 206$\times$ higher token throughput by skipping 74.8% transmissions with 97.4% vocabulary compression, while maintaining 97.4% accuracy.
arXiv:2505.13148v4 Announce Type: replace
Abstract: This paper employs the method of moments (MOM) to calculate the capacitances of a cube and a hollow cylinder. For the cube, each face was divided into a maximum of 600 x 600 sub-areas. By fully exploiting the geometric symmetry between sub-areas and incorporating parallel computing, computational resources were significantly conserved. Our results show that the calculated capacitance of the cube first increases and then decreases as the number of sub-areas increases. When each face was divided into 90 x 90 sub-areas, the capacitance of the unit cube (with an edge length of 1 m) reached a maximum reference value of 73.519014 pF. This indicates that higher accuracy cannot be achieved merely by indefinitely increasing the number of discretized sub-areas. Subsequently, the method was applied to compute the capacitance of a hollow cylinder. The results were compared with numerical solutions based on Lekner's theoretical formula and Cavendish's experimental values, showing good agreement among the three.
arXiv:2505.13878v3 Announce Type: replace
Abstract: Model fusion combines multiple Large Language Models (LLMs) with different strengths into a more powerful, integrated model through lightweight training methods. Existing works on model fusion focus primarily on supervised fine-tuning (SFT), leaving preference alignment (PA) --a critical phase for enhancing LLM performance--largely unexplored. The current few fusion methods on PA phase, like WRPO, simplify the process by utilizing only response outputs from source models while discarding their probability information. To address this limitation, we propose InfiFPO, a preference optimization method for implicit model fusion. InfiFPO replaces the reference model in Direct Preference Optimization (DPO) with a fused source model that synthesizes multi-source probabilities at the sequence level, circumventing complex vocabulary alignment challenges in previous works and meanwhile maintaining the probability information. By introducing probability clipping and max-margin fusion strategies, InfiFPO enables the pivot model to align with human preferences while effectively distilling knowledge from source models. Comprehensive experiments on 11 widely-used benchmarks demonstrate that InfiFPO consistently outperforms existing model fusion and preference optimization methods. When using Phi-4 as the pivot model, InfiFPO improve its average performance from 79.95 to 83.33 on 11 benchmarks, significantly improving its capabilities in mathematics, coding, and reasoning tasks.
arXiv:2605.24760v1 Announce Type: new
Abstract: We present a kinematic and transmission-aware design framework for a serial spherical mechanism with an additional translational degree of freedom for microsurgery. The first contribution is an analytical workspace formulation that provides geometric insight into reachable motion and enables rapid selection of rotation axis orientations without numerical optimization. The second contribution is a dynamics-informed methodology for mechanisms driven by self-locking transmissions, supporting evaluation of torque requirements for a prescribed workspace geometry. The framework is accompanied by an open-source software package for friction identification and inverse dynamics analysis. Experiments on a purpose-built robotic tool for vitreoretinal surgery validate the predictive capability of the models and demonstrate their practical utility for engineering design.
arXiv:2506.17629v2 Announce Type: replace
Abstract: Embodied Visual Reasoning (EVR) seeks to follow complex, free-form instructions based on egocentric video, enabling semantic understanding and spatiotemporal reasoning in dynamic environments. Despite its promising potential, EVR encounters significant challenges stemming from the diversity of complex instructions and the intricate spatiotemporal dynamics in long-term egocentric videos. Prior solutions either employ Large Language Models (LLMs) over static video captions, which often omit critical visual details, or rely on end-to-end Vision-Language Models (VLMs) that struggle with stepwise compositional reasoning. Consider the complementary strengths of LLMs in reasoning and VLMs in perception, we propose CLiViS. It is a novel training-free framework that leverages LLMs for high-level task planning and orchestrates VLM-driven open-world visual perception to iteratively update the scene context. Building on this synergy, the core of CLiViS is a dynamic Cognitive Map that evolves throughout the reasoning process. This map constructs a structured representation of the embodied scene, bridging low-level perception and high-level reasoning. Extensive experiments across multiple benchmarks demonstrate the effectiveness and generality of CLiViS, especially in handling long-term visual dependencies. Code is available at https://github.com/Teacher-Tom/CLiViS.