arXiv:2607.09218v2 Announce Type: replace
Abstract: Whole-arm manipulation involves direct contact with the environment while the robot completes a task by distributing contact across multiple links as contacts form, slide, and break. This setting breaks common implicit assumptions in many learning-based manipulation pipelines: arm configuration tightly couples motion and contact forces, contact state is partially observed under occlusion, and purely learned rollouts can become physically inconsistent under distribution shift because many multi-link contact configurations are sparsely represented in the data. To address this, we propose TACTIC (Tactile and Vision Conditioned Contact-Centric Control), a receding-horizon controller for whole-arm manipulation. TACTIC uses a contact-centric hybrid predictive model that combines RGB-D, distributed tactile sensing, and a compact 2D proximity representation. The model couples a learned, action-conditioned latent dynamics model with analytical kinematics through contact Jacobians, enabling rollouts of future contact configurations and interaction forces. TACTIC integrates these rollouts into a sampling-based MPC planner with contact-aware action sampling: contact Jacobian-based projections steer sampled action sequences toward force-modulating directions, and objectives defined over predicted proximity and interaction forces trade task progress against whole-arm force regulation. We evaluate TACTIC in simulation against state-of-the-art model-based and model-free methods, and perform ablations that isolate the contribution of each design choice. TACTIC consistently outperforms other methods. We further demonstrate real-world performance on a robot with distributed tactile sensing across three whole-arm manipulation tasks that require multi-contact trajectories: turning over and repositioning a manikin, and goal-reaching in a 3D dynamic maze. Website: https://emprise.cs.cornell.edu/tactic
Science Journals
arXiv:2607.06763v2 Announce Type: replace
Abstract: Speculative decoding greatly increases the interactivity of autoregressive language models by trading off computation for extra tokens generated in a single forward pass. Factorized draft models are especially efficient because they predict future-token marginals in parallel, but their independence assumption causes acceptance rates to degrade sharply as the speculative budget grows. We analyze this limitation and introduce Weaver, a lightweight autoregressive adapter that constructs proposal trees from the top-K marginals of a factorized drafter. Weaver restores conditional dependencies between proposed tokens while avoiding a full-vocabulary projection. To support fast verification for models with Gated Delta Net layers, we derive a rollback-free tree-verification algorithm and implement optimized CUDA kernels in SGLang. By combining these model and systems contributions we achieve a 4.37-fold speedup over autoregressive decoding, and outperform a highly optimized DFlash baseline by 24.7%.
arXiv:2607.09629v2 Announce Type: replace
Abstract: Reliable autonomous driving requires full-scene perception that couples foreground objects with dense semantic layout. Recently, 4D millimeter-wave radar has emerged as a robust and affordable sensor, yet its sparse returns make radar-camera fusion necessary for comprehensive scene understanding. Existing radar-camera methods mainly optimize detection, while dual-task systems usually decode boxes and occupancy with limited interaction. To address this gap and advance radar-based multi-task learning, we propose \method, a 4D radar-camera framework for 360$^\circ$ full-scene perception, which models semantic occupancy as a persistent scene state rather than a terminal output. \method{} follows a cross-modal state reasoning paradigm, where the occupancy state is modeled and propagated through stages for coarse-to-fine feature aggregation. Specifically, State-guided BEV Enhancement (SBE) strengthens intra-frame BEV representation, while Doppler-guided Temporal Fusion (DTF) preserves state evidence over longer temporal horizons. Beyond the model, we further extend ManTruckScenes with satellite-map-based generated occupancy labels and pair it with OmniHD-Scenes in a unified cross-dataset detection-and-occupancy protocol. The resulting experiments cover accuracy, robustness, ablation, and efficiency under one radar-camera multi-task evaluation framework. Code and labels will be released upon acceptance.
arXiv:2508.03636v3 Announce Type: replace-cross
Abstract: We propose a Likelihood Matching approach for training diffusion models by first establishing an equivalence between the likelihood of the target data distribution and a likelihood along the sample path of the reverse diffusion. To efficiently compute the reverse sample likelihood, a quasi-likelihood is considered to approximate each reverse transition density by a Gaussian distribution with matched conditional mean and covariance, respectively. The score and Hessian functions for the diffusion generation are estimated by maximizing the quasi-likelihood, ensuring a consistent matching of both the first two transitional moments between every two time points. A stochastic sampler is introduced to facilitate computation that leverages both the estimated score and Hessian information. We establish consistency of the quasi-maximum likelihood estimation, and provide non-asymptotic convergence guarantees for the proposed sampler, quantifying the rates of the approximation errors due to the score and Hessian estimation, dimensionality, and the number of diffusion steps. Empirical and simulation evaluations demonstrate the effectiveness of the proposed Likelihood Matching and validate the theoretical results.
arXiv:2607.10341v1 Announce Type: cross
Abstract: AI visibility measurement is comparative: practitioners want to know which domains generative search engines cite most often and whether observed differences are large enough to support decisions. Yet the industry lacks a principled way to determine whether enough data has been collected. Collection budgets vary widely across studies and platforms, and conclusions are often drawn from rankings whose stability and precision are unknown. We introduce a sequential convergence framework based on two complementary criteria: rank stability evaluates whether the rank-correlation trajectory has reached a structural plateau, while structural sufficiency evaluates whether the spread of citation shares among established domains -- those whose confidence intervals exclude zero -- exceeds the uncertainty of those estimates. Together, these criteria distinguish rankings that have merely stabilized from those sufficiently resolved to support inference. Both are derived from regularities in the observed citation distribution, including its rank structure, uncertainty profile, and the boundary between observed and established domains. The framework retains a small number of structural constants but requires no externally specified query count, correlation target, or confidence-interval width target; stopping is driven by observed measurement uncertainty and remains robust across a range of sufficiency thresholds. Applied across 30 platform-topic combinations spanning Gemini, SearchGPT, and Perplexity, the framework adapts to platform- and topic-specific citation distributions. Results show that no fixed collection budget can be justified across contexts and that convergence can instead be evaluated from the structure of the observed distribution. The framework provides a practical basis for determining when AI visibility measurements are ready to support comparative analysis.
arXiv:2510.05425v2 Announce Type: replace
Abstract: Work environments are often inadequate and lack inclusivity for individuals with upper-body disabilities. This paper presents a novel online framework for adaptive human-robot interaction (HRI) that accommodates users' arm mobility impairments, ultimately aiming to promote active work participation. Unlike traditional human-robot collaboration approaches that assume able-bodied users, our method integrates a mobility model for specific joint limitations into a hierarchical optimal controller. This allows the robot to generate reactive, mobility-aware behaviour online and guides the user's impaired limb to exploit residual functional mobility. The framework was tested in handover tasks involving different upper-limb mobility impairments (i.e., emulated elbow and shoulder arthritis, and wrist blockage), under both standing and seated configurations with task constraints using a mobile manipulator, and complemented by quantitative and qualitative comparisons with state-of-the-art ergonomic HRI approaches. Preliminary results indicated that the framework can personalise the interaction to fit within the user's impaired range of motion and encourage joint usage based on the severity of their functional limitations.
arXiv:2607.09822v1 Announce Type: new
Abstract: Recognition tells an agent what is in an image; personal memory affects what is worth looking up next. In a camera-first setting the user can send only an image, so the agent must form the lookups. We study whether personal visual memory improves agent-side tool choice and tool arguments, and thereby more user-aligned multi-tool lookups. The design uses a three-layer personal visual memory (profile, short-term focus, observations) that is loaded on each turn to condition an LLM tool-calling loop under camera-first intake, and includes conflict-aware write-back intended to refresh the user model for later captures. On 800 images paired with synthetic memory blocks constructed for controlled ablation, removing the full three-layer memory block reduces tool-query relevance by 0.47 points absolute (4.21 -> 3.74 on a 5-point scale; 11.2% relative) and end-to-end utility by 0.082 absolute (0.842 -> 0.760; 9.7% relative). These results measure memory conditioning of tool policy under image-only intake with fixed synthetic blocks, not multi-session write-back from live user histories.
arXiv:2607.10349v1 Announce Type: new
Abstract: We present a train-resolved SAXS methodology for recovering weak scattering signals from high-repetition-rate XFEL datasets and apply it to aqueous L-cysteine solutions measured at the European XFEL. Independent scale-plus-offset fitting was performed for matched cysteine and water train pairs, followed by subtraction of transmission-matched water--water controls. The 0.5 M dataset reveals a reproducible sign-changing residual SAXS signal that increases with incident XFEL transmission and remains after removal of detector-wide scaling, additive offsets, and matched water--water control residuals. Convergence and block-averaging analyses show that the residual emerges progressively as independent train pairs are accumulated and exhibits uncertainty scaling close to the expected inverse square-root dependence on N. These results establish a statistically robust transmission-dependent residual SAXS contribution whose microscopic origin remains unresolved, while demonstrating that train-resolved observables combined with matched controls can substantially improve sensitivity to weak scattering signals in high-repetition-rate XFEL experiments.
arXiv:2510.07510v2 Announce Type: replace-cross
Abstract: Optically-active spin qubits have emerged as powerful quantum sensors capable of nanoscale magnetometry, yet conventional coherent sensing approaches are ultimately limited by the coherence time of the sensor, typically precluding detection in the sub-MHz regime. We present a broadly applicable fluorescence-encoding method that circumvents coherence-time constraints by transducing time-varying magnetic fields directly into modulated fluorescence signals. Using nitrogen-vacancy centers in diamond as a model system, we demonstrate shot-noise-limited sensitivity for AC magnetic fields spanning near-DC to MHz frequencies, with detection bandwidth tunable via optical excitation power. The technique captures complete spectral information in a single measurement, eliminating the need for point-by-point frequency scanning, and allows phase-sensitive multi-frequency detection with Hz-level resolution. This approach transforms quantum sensors into atomic-scale spectrum analyzers, with immediate applications for low-frequency RF communication, zero-field NMR, and bioelectronic sensing. Our approach is broadly applicable to the expanding class of optically-active spin qubits, including molecular systems and fluorescent proteins, opening new sensing regimes previously inaccessible to coherent techniques
arXiv:2510.11546v3 Announce Type: replace-cross
Abstract: High-dimensional regression often suffers from heavy-tailed noise and outliers, which can severely undermine the reliability of least-squares based methods. To improve robustness, we adopt a non-smooth Wilcoxon score based rank objective and incorporate the group sparsity regularization. By extending the tuning-free property originally developed for the rank Lasso, we introduce a simulation-based tuning rule and further establish a finite-sample error bound for the resulting estimator. To solve the associated optimization problem, we develop a proximal augmented Lagrangian method, for which we provide a novel convergence analysis by proving the metric subregularity of the underlying non-polyhedral KKT mapping, while enabling efficient semismooth Newton updates for the subproblems. Extensive numerical experiments demonstrate the robustness and effectiveness of our proposed estimator against several leading alternatives, and showcase the efficiency and scalability of our algorithm compared to the state-of-the-art baseline in both simulated and real-data settings.
arXiv:2510.21354v2 Announce Type: replace-cross
Abstract: Through systematic experimental and theoretical studies of layer-thickness-dependent absorption in semiconducting MoSe$_2$ and WS$_2$ across the visible to near-infrared spectral range, we demonstrate a universal absorption behavior in solids at nanoscale thicknesses. With increasing thickness, a non-monotonic evolution of absorption integrated over the measured spectral region is revealed which is accompanied by pronounced oscillatory features. This shows a strong deviation from the expected Beer-Lambert law. Below 10 nm, we observe a sharp anomalous increase in absorption, with deviations from Beer's law exceeding 50% in layered semiconductors. Our conclusions hold irrespective of the presence of any optical resonances such as excitons or plasmons within the spectral window. The observed behavior has origins in the electromagnetic interference effects taking place between the two surfaces of the thin crystals. The present work on 2D semiconductors is extendable to all kinds of solids such as conventional semiconductors (e.g. Si, GaAs, GaN, InP), (semi) metals (e.g. Al, Ag, Au, c-HOPG) and 2D magnetic materials (e.g. CrSBr and NiPS$_3$). Our results provide fundamental insights into light-matter interactions in solids at the nanoscale and are vital for optimally designing the new generation of absorption-based flexible optoelectronic devices.
arXiv:2604.10614v2 Announce Type: replace-cross
Abstract: We introduce new kinetic equations to describe epidemics' spread while accounting for individuals' opinions on protective behaviors. Opinion exchanges occur on a social network represented by a graphon, whose choice strongly influences the dynamics and leads to the emergence of complex nonlinear phenomena, like the creation of opinion leaders or the spontaneous formation of epidemic waves. Starting from individual-based interactions, we derive a nonlinear nonlocal Fokker-Planck model involving reaction terms and degenerate drift-diffusion operators, which depend on the underlying graphon. We establish rigorous results of convergence to equilibrium in $L^1$ space, via relative entropy estimates, and in homogeneous Sobolev spaces $\dot{H}^{-s}$, $s \in \big(\frac{1}{2}, 1\big)$, using Fourier-based techniques. We then design a structure-preserving scheme for the coupled opinion-epidemiological system, highlighting graphon effects: opinion leaders supporting protective behaviors limit disease spread, whereas influenceable individuals may shift toward opposing views, worsening epidemics. At last, we introduce a time-dependent quantity analogous to the effective reproduction number, whose oscillations are linked with the formation of epidemic waves. Notably, these waves are not induced by an explicit external forcing but they naturally emerge from the interactions between agents, depending on the connectivity level prescribed by the graphon.
arXiv:2607.10365v1 Announce Type: new
Abstract: Vision Transformers (ViTs) are difficult to interpret because current methods of relevance propagation and attention flow do not fully consider some key architectural features, such as the uneven importance of attention heads and residual connections. Prior approaches typically assume uniform importance across attention heads; furthermore, they model skip connections as identity paths, leading to inaccurate relevance attribution. To address these issues, we introduce GradSkip, a novel relevance propagation method for ViTs based on adaptive head weighting and skip-aware propagation. GradSkip models the different importance of the attention heads and dynamically distributes relevance between the attention and residual paths. Experiments on ImageNet1K and BloodMNIST demonstrate a state-of-the-art faithfulness of GradSkip while requiring over 14 times fewer GFLOPs than the best-performing existing approaches. Additional evaluations using transformer-based segmentation confirm improved localization and alignment with ground-truth regions.
arXiv:2607.11457v1 Announce Type: new
Abstract: Many forecast applications require high frequency temporal resolution, yet most state-of-the-art data-driven weather forecasting systems operate at 6-hourly resolution. Although direct hourly forecasting is possible, it suffers from error accumulation and temporal inconsistency. We introduce HourGlass, a probabilistic data-driven temporal downscaling method that reconstructs the evolution between forecast states. HourGlass is trained using variants of the continuous ranked probability score (CRPS) preserving small-scale spatial variability while encouraging temporal consistency. Unlike existing deterministic temporal downscaling approaches, which tend to produce overly smooth fields, HourGlass generates realistic probabilistic forecasts. Training on forecast trajectories rather than reanalysis or analysis data also avoids the temporal inconsistencies present in datasets used by previous methods. We evaluate HourGlass in two settings: AIFS-HourGlass, applied globally to ECMWF's AIFS-Single and AIFS-ENS forecast systems, and Bris-HourGlass, applied regionally to MET Norway's high-resolution stretched-grid ensemble model, Bris. Verification against observations shows that both models retain the skill of their underlying forecasting systems while producing temporally coherent hourly forecasts with realistic small-scale variability. Case studies demonstrate physically consistent evolution during rapidly developing weather events, including extratropical cyclones and organised convection. Hourly precipitation remains challenging: HourGlass improves the spatial realism of precipitation fields but still underestimates the most intense extremes, a common limitation of data-driven weather forecasting models. These results demonstrate that HourGlass effectively bridges the gap between 6-hourly data-driven forecasts and the hourly products required for operational regional and global forecasting.
arXiv:2604.13213v3 Announce Type: replace-cross
Abstract: Rare events such as conformational changes in biomolecules, phase transitions, and chemical reactions are central to the behavior of many physical systems, yet they are extremely difficult to study computationally because unbiased simulations seldom produce them. Transition Path Theory (TPT) provides a rigorous statistical framework for analyzing such events: it characterizes the ensemble of reactive trajectories between two designated metastable states (reactant and product), and its central object--the committor function, which gives the probability that the system will next reach the product rather than the reactant--encodes all essential kinetic and thermodynamic information. We introduce a framework that casts committor estimation as a stochastic optimal control (SOC) problem. In this formulation the committor defines a feedback control--proportional to the gradient of its logarithm--that actively steers trajectories toward the reactive region, thereby enabling efficient sampling of reactive paths. To solve the resulting hitting-time control problem we develop two complementary objectives: a direct backpropagation loss and a principled off-policy Value Matching loss, for which we establish first-order optimality guarantees. We further address metastability, which can trap controlled trajectories in intermediate basins, by introducing an alternative sampling process that preserves the reactive current while lowering effective energy barriers. On benchmark systems, the framework yields markedly more accurate committor estimates, reaction rates, and equilibrium constants than existing methods.
arXiv:2606.07852v2 Announce Type: replace-cross
Abstract: Unambiguous state discrimination (USD) measurements are attractive because outcomes are either marked as conclusive (i.e., error free) or inconclusive (i.e., erased). We study affine filtering measurements, a structured variant of USD for decoding classical linear codes over pure-state classical-quantum channels, where a conclusive outcome identifies an affine subspace containing the transmitted codeword and an inconclusive outcome is treated as an erasure. For a group-covariant indexing of pure-state codewords, we show that the optimal design of affine filtering measurements is a semidefinite program that can be reduced to a linear program via character-based diagonalization. We use the resulting measurement to build a quantum decoding framework for local codes, and we demonstrate (via simulations on regular LDPC codes from Gallager ensembles using single parity check local constraints) that affine filtering based decoding can outperform symbol-wise USD and symbol-wise pretty good measurement based decoding methods on i.i.d. pure-state channels. In an independent and concurrent work, Buzet and Chailloux study similar fine-grained USD measurements for symmetric families of states. Their focus is on the code-agnostic setting whereas our focus is on code-aware constructions and decoding.
arXiv:2607.11771v1 Announce Type: new
Abstract: Explainability has emerged as a critical requirement for AI-based systems, particularly in safety-critical and regulated domains. Although prior research has proposed frameworks, patterns, and user-centered approaches to support explainability, there is limited empirical understanding of how existing Requirements Engineering (RE) practices support explainability requirements across the RE lifecycle, especially in an industrial context. This paper reports early findings from an ongoing industry-based study investigating how explainability requirements are elicited, specified, and validated using established RE techniques. We conducted a multi-phase qualitative study with eight practitioners at Daimler Truck, employing think-aloud protocols and moderated group discussions across requirements elicitation, specification, and validation steps. Our preliminary analysis reveals recurring challenges across all steps, including conceptual ambiguity during elicitation, limited testability and expressiveness during specification, and fragmented validation due to vague criteria and regulatory uncertainty. These findings indicate that current RE practices provide limited support to systematically address explainability requirements. The paper contributes empirical insights into step-specific and cross-cutting challenges and outlines a research vision toward developing an empirically grounded RE framework for explainable AI-based systems.
arXiv:2501.12982v3 Announce Type: replace-cross
Abstract: This paper investigates how diffusion generative models leverage (unknown) low-dimensional structure to accelerate sampling. Focusing on two mainstream samplers -- the denoising diffusion implicit model (DDIM) and the denoising diffusion probabilistic model (DDPM), we prove that their iteration complexities under exact score functions are at most the order of $k/\varepsilon$ (up to log factor), where $\varepsilon$ is the precision in total variation distance and $k$ is some intrinsic dimension of the target distribution. We further extend these convergence guarantees to the setting in which the score functions are learned from data rather than known exactly, showing that the convergence performance degrades gracefully under suitable score estimation assumptions. We then show that these assumptions are attainable via kernel-based score estimators with finite-sample guarantees that also adapt to the low-dimensional structure. Our results apply to a broad family of target distributions without requiring smoothness or log-concavity. Our findings provide the first rigorous evidence for the adaptivity of the DDIM-type samplers to unknown low-dimensional structure, and improve over the state-of-the-art DDPM theory regarding total variation convergence.
arXiv:2606.29189v2 Announce Type: replace-cross
Abstract: We report the first continuous families of stable, ballistic, prograde cycler orbits in the circular restricted three-body problem: periodic trajectories that alternately undergo temporary capture and orbit each primary. We construct continuous families of symmetric cyclers from intersections of the stable and unstable manifold tubes of the $L_1$ Lyapunov orbit and exhibit stable examples across more than two orders of magnitude in mass ratio, from the Sun--Jupiter regime to the equal-mass limit. Linear stability separates naturally into planar and out-of-plane components. A planar-stable branch of every computed family is created simultaneously with a hyperbolic branch in a saddle-center bifurcation of the return map at the family's maximal Jacobi constant, while out-of-plane instability occurs only through isolated parametric resonances. Every computed family contains a subfamily that is linearly stable to both planar and out-of-plane perturbations. These stable cyclers persist into the full three-body problem for sufficiently small third mass. We conjecture that saddle-center birth is universal among cycler families, suggesting that stable cyclers are a generic feature of three-body dynamics.
arXiv:2607.10656v1 Announce Type: cross
Abstract: Characterizing quantum topological phases requires measuring non-local string order parameters, demanding access to the full system, which is often experimentally unfeasible. In this work, we introduce a data-efficient supervised learning framework that circumvents this limitation by recognizing quantum phases from small subsystems. Our protocol utilizes a quantum kernel constructed from the reduced density matrices of these subsystems, which can be efficiently estimated experimentally. We benchmark our framework with the classification of the phase diagrams of two spin models on one-dimensional lattices, namely the generalized cluster-Ising spin-1/2 chain and the anisotropic Haldane spin-1 chain. Remarkably, our approach achieves high accuracy in phase classification when operations are limited to as few as one to four sites, and it also generalizes to longer chains even when trained on moderate system sizes. These findings demonstrate that local reduced density matrices preserve vital signatures of global topological phases, offering a practical route to characterize rich phase diagrams of quantum many-body systems.
arXiv:2607.10664v1 Announce Type: cross
Abstract: In this paper, we provide a systematic investigation of SO(2) theory to machine learning interatomic potentials (MLIPs) and identify the limitations of conventional SO(2) Linear architectures relative to SO(3) Clebsch-Gordan Tensor Products (CGTP). Building on these insights, we propose direct Cartesian construction and recursive Clebsch-Gordan construction of Wigner D-matrices and introduce two novel interaction building blocks. First, we propose the Edge Complex Product Basis based on Generalized Asymmetric Contraction, a new formulation for many-body expansion that directly constructs higher-order interactions on edges through complex-valued equivariant multiplications. Second, we introduce Radial Rotary Complex Attention(RRA), which enhances extrapolation performance and surpasses existing attention vector formulations. We also introduce several improvements to the Atomic Cluster Expansion module. Building on these advances, we train our models on OMat24, sAlex, and MPTrj, and introduce TECE-OAM-RRA-1.0, which achieve state-of-the-art (SOTA) performance on the Matbench Discovery.
arXiv:2602.22861v3 Announce Type: replace
Abstract: We develop structure-preserving discontinuous Galerkin methods for the Cahn-Hilliard-Navier-Stokes equations with degenerate mobility. The proposed SWIPD-L and SIPGD-L methods incorporate parametrized mobility fluxes with edge-wise mobility treatments for enhanced coercivity-stability control. We prove coercivity for the generalized trilinear form and demonstrate optimal convergence rates while preserving mass conservation, energy dissipation, and the discrete maximum principle. Comparisons with existing SIPG-L and SWIP-L methods confirm similar stability. Validation on $hp$-adaptive meshes for both standalone Cahn-Hilliard and coupled systems shows significant computational savings without accuracy loss.
arXiv:2607.05567v2 Announce Type: replace-cross
Abstract: The flip graph of an origami crease pattern has the flat-foldable mountain-valley assignments as vertices, and an edge joins two of them that differ by a single face flip. A basic invariant of this graph is the degree sequence, which counts the vertices of each degree. On the $m\times n$ Miura-ori, this sequence is known as a bivariate polynomial only for small degrees, each count obtained by a separate argument. This paper gives one uniform construction that expresses, for every degree $d$, the number of degree-$d$ vertices as a single symmetric polynomial in $(m,n)$ for all sufficiently large $m,n$. Subject to a single degree bound, this polynomial has total degree $d-2$, growing for $d\ge5$ as an explicit multiple of $m^{d-2}+n^{d-2}$; the bound is proved here when the count splits into independent row and column factors, and open otherwise. The region is $m,n\ge\max(d-1,2)$; the polynomials are computed in closed form through $d=10$, and the bound is verified in every case through $d=7$. Below this region, the count departs from the polynomial by a correction whose leading coefficient, through degree eleven, is $-4$ times a Baxter number. Each such polynomial thus counts the Miura-ori's flat-foldable assignments admitting exactly $d$ single face flips.
arXiv:2607.11526v1 Announce Type: new
Abstract: Material property prediction (MPP) infers key properties from chemical composition and structure, accelerating the discovery and optimization of novel materials. In the realm of MPP, MatBench is a widely accepted benchmarking tool that defines over ten significant problems and provides the paradigm of performance evaluation for AI prediction models. Even though MatBench works well in benchmarking the performances of prediction models on in-distribution (ID) tasks and datasets, it lacks the ability to reflect their performances on out-of-distribution (OOD) material data, resulting failure in new material discovery. By combining the pipelines of MatBench and the existing researches on OOD performance evaluation, this study enables a huge space of benchmarking configurations, comprehensively reflecting the performances, abilities, and disadvantages of various AI prediction models. This work reports that the discrepancy of performances at different configuration values is huge and can be illustrated with prior knowledge and novel insights, therefore consideration of causal effect of configurations on performance results is necessary. In case of the impossibility of enumerative benchmarking at every configuration, this work further proposes AutoMatBench, an automatic toolkit with Bayesian optimization. Experiments with AutoMatBench reports that, within twelve steps of optimization, the similar results with MatBench and former OOD research can be accessed while more than half of the cost are saved. Besides, this tool also yields more essential findings on MPP benchmarking, positively contributing to the cost and efficiency of new material discovery.
arXiv:2607.11529v1 Announce Type: new
Abstract: In this paper, we tackle the Aerial Vision-and-Dialog Navigation (AVDN) task in the training-free setting for resource-efficient high-altitude UAV navigation.Naively applying MLLMs leads to unreliable navigation due to weak directional grounding and the lack of explicit spatial memory.To address these issues, we propose PSC-AVDN, a training-free framework that tightly couples a three-stage Parsing-Search-Confirmation reasoning pipeline with a Structured Spatial Memory (SSM).The parsing stage uses an LLM to convert ambiguous dialogue instructions into stable geometric directional and destination cues.A Search Chain-of-Thought (S-CoT) then performs stepwise target exploration under high-altitude observations, and a Confirmation Chain-of-Thought (C-CoT) conducts fine-grained verification around candidate regions to resolve visual ambiguity.Meanwhile, SSM integrates three complementary sources of spatial cues, including multi-scale visual observation, spatial visual memory, and structured geometric memory to provide global spatial context and long-horizon consistency.Extensive experiments on ANDH and ANDH-Full show that PSC-AVDN establishes new state-of-the-art performance in the training-free setting, matching or surpassing several finetuned methods.Code will be publicly available at: https://github.com/QY6616/PSC-AVDN