High-density through-substrate vias (TSVs) enable 2.5D/3D heterogeneous integration but introduce significant signal-integrity and thermal-reliability challenges due to electrical coupling, insertion loss, and self-heating. Conventional full-wave finite-element method (FEM) simulations provide high accuracy but become computationally prohibitive for large design-space exploration. This work presents a scalable electro-thermal modeling and optimization framework that combines physics-informed analytical modeling, graph neural network (GNN) surrogates, and full-wave sign-off validation. A multi-conductor analytical model computes broadband S-parameters and effective anisotropic thermal conductivities of TSV arrays, achieving $5\%-10\%$ relative Frobenius error (RFE) across array sizes up to $15x15$. A physics-informed GNN surrogate (TSV-PhGNN), trained on analytical data and fine-tuned with HFSS simulations, generalizes to larger arrays with RFE below $2\%$ and nearly constant variance. The surrogate is integrated into a multi-objective Pareto optimization framework targeting reflection coefficient, insertion loss, worst-case crosstalk (NEXT/FEXT), and effective thermal conductivity. Millions of TSV configurations can be explored within minutes, enabling exhaustive layout and geometric optimization that would be infeasible using FEM alone. Final designs are validated with Ansys HFSS and Mechanical, showing strong agreement. The proposed framework enables rapid electro-thermal co-design of TSV arrays while reducing per-design evaluation time by more than six orders of magnitude.

Large Audio Language Models (LALMs) achieve strong performance on audio-language tasks; however, their reliability in real-world settings remains underexplored. We introduce Audio Hallucination Attacks (AHA), an attack suite called AHA-Eval, comprising 6.5K QA pairs designed to test whether LALMs genuinely ground their responses in the audio input. AHA targets two attack surfaces: (i) query-based attacks, which exploit question structure to induce hallucinations about absent sounds, and (ii) audio-based attacks, which inject synthetic speech describing non-existent events into the audio stream. Evaluating state-of-the-art LALMs, including Audio Flamingo 3 and Gemini 3 Pro, we observe high attack success rates of 95.35% and 79.65%, respectively, revealing a reliability gap that is hidden by standard benchmark performance. To mitigate this, we propose a 120K QA post-alignment dataset, AHA-Guard, which successfully reduces attack success rates by up to 49%.

Grokking in modular arithmetic has established itself as the quintessential fruit fly experiment, serving as a critical domain for investigating the mechanistic origins of model generalization. Despite its significance, existing research remains narrowly focused on specific local circuits or optimization tuning, largely overlooking the global structural evolution that fundamentally drives this phenomenon. We propose that grokking originates from a spontaneous simplification of internal model structures governed by the principle of parsimony. We integrate causal, spectral, and algorithmic complexity measures alongside Singular Learning Theory to reveal that the transition from memorization to generalization corresponds to the physical collapse of redundant manifolds and deep information compression, offering a novel perspective for understanding the mechanisms of model overfitting and generalization.

Vision-Language Models (VLMs) have demonstrated strong capabilities across a wide range of multimodal tasks. However, recent studies have shown that VLMs, such as CLIP, perform poorly in understanding negation expressions, which are common in natural language. In this work, we propose Omni-NegCLIP, a fine-tuned CLIP model that improves CLIP's understanding of two types of negation, namely presence-based negation and absence-based negation, which correspond to negated expressions of objects that are actually present in an image and those that may plausibly exist in an image but are in fact absent, respectively, by modifying CLIP's original InfoNCE contrastive loss. Specifically, we design a presence-based contrastive objective that pulls image embeddings closer to their original caption embeddings while pushing them away from the corresponding presence-based negated caption embeddings, and an absence-based contrastive objective that aligns image embeddings with both original and absence-based negated caption embeddings while maintaining a semantic distinction between the two text embeddings. Based on our observation that the front transformer layers of CLIP text encoder have stronger learning ability for negated text than the later layers, we fine-tune the front transformer layers of the CLIP text encoder at each training step using the combined contrastive objective. Experimental results show that, compared with pretrained CLIP, Omni-NegCLIP improves performance on presence-based negation and absence-based negation tasks by up to 52.65% and 12.50%, respectively, without sacrificing general capability in image-text retrieval and even improving it by up to 19.62%. Compared with prior works, Omni-NegCLIP demonstrates a more comprehensive ability to understand multiple types of negation tasks.

Long video understanding is a key challenge that plagues the advancement of \emph{Multimodal Large language Models} (MLLMs). In this paper, we study this problem from the perspective of visual memory mechanism, and proposed a novel and training-free approach, termed \emph{Flexible Memory} (\textbf{FlexMem}). In principle, FlexMem aims to mimic human behavior of video watching, \emph{i.e.}, continually watching video content and recalling the most relevant memory fragments to answer the question. In this way, FlexMem can help MLLMs achieve video understanding of infinite lengths, unlike previous methods that process all video information at once and have input upper-limit. Concretely, FlexMem first consider the visual KV caches as the memory sources, and realize the effective memory transfer and writing via a dual-pathway compression design. Afterwards, FlexMem also explores different memory reading strategies for the diverse video understanding tasks, including the popular streaming one. To validate FlexMem, we apply it to two popular video-MLLMs, and conduct extensive experiments on five long video and one streaming video task. The experimental results show that on \textbf{a single 3090 GPU}, our FlexMem can achieve obvious improvements than existing efficient video understanding methods and process more than \textbf{1k frames}, which also helps the base MLLMs achieve comparable or even better performance than SOTA MLLMs on some benchmarks, \emph{e.g.} , GPT-4o and Gemini-1.5 Pro.

Commercial lunar activity is accelerating the need for reliable surface infrastructure and routine operations to keep it functioning. Maintenance tasks such as inspection, cleaning, dust mitigation, and minor repair are essential to preserve performance and extend system life. A specific application is the cleaning of lunar solar arrays. Solar arrays are expected to provide substantial fraction of lunar surface power and operate for months to years, supplying continuous energy to landers, habitats, and surface assets, making sustained output mission-critical. However, over time lunar dust accumulates on these large solar arrays, which can rapidly degrade panel output and reduce mission lifetime. We propose a small mobile robot equipped with a long-reach, lightweight deployable boom and interchangeable cleaning tool to perform gentle cleaning over meter-scale workspaces with minimal human involvement. Building on prior vision-guided long-reach manipulation, we add a compliant wrist with distal force sensing and a velocity-based admittance controller to regulate stable contact during surface cleaning. In preliminary benchtop experiments on a planar surface, the system maintained approximately 2 N normal force while executing a simple cleaning motion over boom lengths from 0.3 m to 1.0 m, with RMS force error of approximately 0.2 N after initial contact. These early results suggest that deployable long-reach manipulators are a promising architecture for robotic maintenance of lunar infrastructure such as solar arrays, radiators, and optical surfaces.

Can a diffusion model produce its own "mental average" of a concept-one that is as sharp and realistic as a typical sample? We introduce Diffusion Mental Averages (DMA), a model-centric answer to this question. While prior methods aim to average image collections, they produce blurry results when applied to diffusion samples from the same prompt. These data-centric techniques operate outside the model, ignoring the generative process. In contrast, DMA averages within the diffusion model's semantic space, as discovered by recent studies. Since this space evolves across timesteps and lacks a direct decoder, we cast averaging as trajectory alignment: optimize multiple noise latents so their denoising trajectories progressively converge toward shared coarse-to-fine semantics, yielding a single sharp prototype. We extend our approach to multimodal concepts (e.g., dogs with many breeds) by clustering samples in semantically-rich spaces such as CLIP and applying Textual Inversion or LoRA to bridge CLIP clusters into diffusion space. This is, to our knowledge, the first approach that delivers consistent, realistic averages, even for abstract concepts, serving as a concrete visual summary and a lens into model biases and concept representation.

The ski rental problem is a canonical model for online decision-making under uncertainty, capturing the fundamental trade-off between repeated rental costs and a one-time purchase. While classical algorithms focus on worst-case competitive ratios and recent "learning-augmented" methods leverage point-estimate predictions, neither approach fully exploits the richness of full distributional predictions while maintaining rigorous robustness guarantees. We address this gap by establishing a systematic framework that integrates distributional advice of unknown quality into both deterministic and randomized algorithms. For the deterministic setting, we formalize the problem under perfect distributional prediction and derive an efficient algorithm to compute the optimal threshold-buy day. We provide a rigorous performance analysis, identifying sufficient conditions on the predicted distribution under which the expected competitive ratio (ECR) matches the classic optimal randomized bound. To handle imperfect predictions, we propose the Clamp Policy, which restricts the buying threshold to a safe range controlled by a tunable parameter. We show that this policy is both robust, maintaining good performance even with large prediction errors, and consistent, approaching the optimal performance as predictions become accurate. For the randomized setting, we characterize the stopping distribution via a Water-Filling Algorithm, which optimizes expected cost while strictly satisfying robustness constraints. Experimental results across diverse distributions (Gaussian, geometric, and bi-modal) demonstrate that our framework improves consistency significantly over existing point-prediction baselines while maintaining comparable robustness.

Large language models (LLMs) are widely applied to data analytics over documents, yet direct reasoning over long, noisy documents remains brittle and error-prone. Hence, we study document question answering (QA) that consolidates dispersed evidence into a structured output (e.g., a table, graph, or chunks) to support reliable, verifiable QA. We propose a two-pillar framework, LiteCoST, to achieve both high accuracy and low latency with small language models (SLMs). Pillar 1: Chain-of-Structured-Thought (CoST). We introduce a CoST template, a schema-aware instruction that guides a strong LLM to produce both a step-wise CoST trace and the corresponding structured output. The process induces a minimal structure, normalizes entities/units, aligns records, serializes the output, and verifies/refines it, yielding auditable supervision. Pillar 2: SLM fine-tuning. The compact models are trained on LLM-generated CoST data in two stages: Supervised Fine-Tuning for structural alignment, followed by Group Relative Policy Optimization (GRPO) incorporating triple rewards for answer/format quality and process consistency. By distilling structure-first behavior into SLMs, this approach achieves LLM-comparable quality on multi-domain long-document QA using 3B/7B SLMs, while delivering 2-4x lower latency than GPT-4o and DeepSeek-R1 (671B). The code is available at https://github.com/HKUSTDial/LiteCoST.

Existing benchmarks measure capability -- whether a model succeeds on a single attempt -- but production deployments require reliability -- consistent success across repeated attempts on tasks of varying duration. We show these properties diverge systematically as task duration grows, and that pass@1 on short tasks is structurally blind to this divergence. We introduce a reliability science framework for long-horizon LLM agents with four metrics: Reliability Decay Curve (RDC), Variance Amplification Factor (VAF), Graceful Degradation Score (GDS), and Meltdown Onset Point (MOP). We evaluate 10 models across 23,392 episodes on a 396-task benchmark spanning four duration buckets and three domains. Key findings: (1) reliability decay is domain-stratified -- SE GDS drops from 0.90 to 0.44 while document processing is nearly flat (0.74 to 0.71); (2) VAF bifurcates by capability tier -- high VAF is a capability signature, not an instability signal; (3) capability and reliability rankings diverge substantially, with multi-rank inversions at long horizons; (4) frontier models have the highest meltdown rates (up to 19%) because they attempt ambitious multi-step strategies that sometimes spiral; and (5) memory scaffolds universally hurt long-horizon performance across all 10 models. These results motivate reliability as a first-class evaluation dimension alongside capability.

Infrared target detection (IRSTD) tasks have critical applications in areas like wilderness rescue and maritime search. However, detecting infrared targets is challenging due to their low contrast and tendency to blend into complex backgrounds, effectively camouflaging themselves. Additionally, other objects with similar features (distractors) can cause false alarms, further degrading detection performance. To address these issues, we propose a novel \textbf{C}amouflage-aware \textbf{C}ounter-\textbf{D}istraction \textbf{Net}work (CCDNet) in this paper. We design a backbone with Weighted Multi-branch Perceptrons (WMPs), which aggregates self-conditioned multi-level features to accurately represent the target and background. Based on these rich features, we then propose a novel Aggregation-and-Refinement Fusion Neck (ARFN) to refine structures/semantics from shallow/deep features maps, and bidirectionally reconstruct the relations between the targets and the backgrounds, highlighting the targets while suppressing the complex backgrounds to improve detection accuracy. Furthermore, we present a new Contrastive-aided Distractor Discriminator (CaDD), enforcing adaptive similarity computation both locally and globally between the real targets and the backgrounds to more precisely discriminate distractors, so as to reduce the false alarm rate. Extensive experiments on infrared image datasets confirm that CCDNet outperforms other state-of-the-art methods.

Accurate and efficient environment representation is crucial for robotic applications such as motion planning, manipulation, and navigation. Signed distance functions (SDFs) have emerged as a powerful representation for encoding distance to obstacle boundaries, enabling efficient collision-checking and trajectory optimization techniques. However, existing SDF reconstruction methods have limitations when it comes to large-scale uncertainty-aware SDF estimation from streaming sensor data. Voxel-based approaches are limited by fixed resolution and lack uncertainty quantification, neural network methods require significant training time, while Gaussian process (GP) methods struggle with scalability, sign estimation, and uncertainty calibration. In this letter, we develop an open-source library, Kernel-SDF, which uses kernel regression to learn SDF with calibrated uncertainty quantification in real-time. Our approach consists of a front-end that learns a continuous occupancy field via kernel regression, and a back-end that estimates accurate SDF via GP regression using samples from the front-end surface boundaries. Kernel-SDF provides accurate SDF, SDF gradient, SDF uncertainty, and mesh construction in real-time. Evaluation results show that Kernel-SDF achieves superior accuracy compared to existing methods, while maintaining real-time performance, making it suitable for various robotics applications requiring reliable uncertainty-aware geometric information.

Future infrastructure construction on the lunar surface will require semi- or fully-autonomous operation from robots deployed at the build site. In particular, tasks such as electrical outfitting necessitate transport, routing, and fine manipulation of cables across large structures. To address this need, we present a compact and long-reach manipulator incorporating a deployable composite boom, capable of performing manipulation tasks across large structures and workspaces. We characterize the deflection, vibration, and blossoming characteristics inherent to the deployable structure, and present a manipulation control strategy to mitigate these effects. Experiments indicate an average endpoint accuracy error of less than 15 mm for boom lengths up to 1.8 m. We demonstrate the approach with a cable routing task to illustrate the potential for lunar outfitting applications that benefit from long reach.

Fine-scale-faithful neural simulation under fixed storage budgets remains challenging. Many existing methods reduce high-frequency error by improving architectures, training objectives, or rollout strategies. However, under budgeted coarsen-quantize-decode pipelines, fine detail can already be lost when the carried state is constructed. In the canonical periodic incompressible Navier-Stokes setting, we show that primitive and derived fields undergo systematically different retained-band distortions under the same operator. Motivated by this observation, we formulate Derived-Field Optimization (DerivOpt), a general state-design framework that chooses which physical fields are carried and how storage budget is allocated across them under a calibrated channel model. Across the full time-dependent forward subset of PDEBench, DerivOpt not only improves pooled mean rollout nRMSE, but also delivers a decisive advantage in fine-scale fidelity over a broad set of strong baselines. More importantly, the gains are already visible at input time, before rollout learning begins. This indicates that the carried state is often the dominant bottleneck under tight storage budgets. These results suggest a broader conclusion: in budgeted neural simulation, carried-state design should be treated as a first-class design axis alongside architecture, loss, and rollout strategy.

SiPaKosa is a comprehensive corpus of Sinhala and Pali doctrinal texts comprising approximately 786K sentences and 9.25M words, incorporating 16 copyright-cleared historical Buddhist documents alongside the complete web-scraped Tripitaka canonical texts. The corpus was created through high-quality OCR using Google Document AI on historical manuscripts, combined with systematic web scraping of canonical repositories, followed by rigorous quality control and metadata annotation. The corpus is organised into language-specific subcorpora: Sinhala and Mixed Sinhala-Pali. We evaluate the performance of language models using ten pretrained models, with perplexity scores ranging from 1.09 to 189.67 on our corpus. This analysis shows that proprietary models significantly outperform open-source alternatives by factors of three to six times. This corpus supports the pretraining of domain-adapted language models, facilitates historical language analysis, and aids in the development of information retrieval systems for Buddhist scholarship while preserving Sinhala cultural heritage.

Sign language is the primary approach of communication for the Deaf and Hard-of-Hearing (DHH) community. While there are numerous benchmarks for high-resource sign languages, low-resource languages like Arabic remain underrepresented. Currently, there is no publicly available dataset for Syrian Arabic Sign Language (SyArSL). To overcome this gap, we introduce SyriSign, a dataset comprising 1500 video samples across 150 unique lexical signs, designed for text-to-SyArSL translation tasks. This work aims to reduce communication barriers in Syria, as most news are delivered in spoken or written Arabic, which is often inaccessible to the deaf community. We evaluated SyriSign using three deep learning architectures: MotionCLIP for semantic motion generation, T2M-GPT for text-conditioned motion synthesis, and SignCLIP for bilingual embedding alignment. Experimental results indicate that while generative approaches show strong potential for sign representation, the limited dataset size constrains generalization performance. We will release SyriSign publicly, hoping it serves as an initial benchmark.

Dexterous hand teleoperation requires motion re-targeting methods that simultaneously achieve high-frequency real-time performance and enforcement of heterogeneous kinematic and safety constraints. Existing nonlinear optimization-based approaches often incur prohibitive computational cost, limiting their applicability to kilohertz-level control, while learning-based methods typically lack formal safety guarantees. This paper proposes a scalable motion retargeting framework that reformulates the nonlinear retargeting problem into a convex quadratic program in joint differential space. Heterogeneous constraints, including kinematic limits and collision avoidance, are incorporated through systematic linearization, resulting in improved computational efficiency and numerical stability. Control barrier functions are further integrated to provide formal safety guarantees during the retargeting process. The proposed framework is validated through simulations and hardware experiments on the Wuji Hand platform, outperforming state-of-the-art methods such as Dex-Retargeting and GeoRT. The framework achieves high-frequency operation with an average latency of 9.05 ms, while over 95% of retargeted frames satisfy the safety criteria, effectively mitigating self-collision and penetration during complex manipulation tasks.

In recent years, multimodal large models have continued to improve on general benchmarks. However, in real-world content moderation and adversarial settings, mainstream models still suffer from degraded generalization and catastrophic forgetting because of limited fine-grained visual perception and insufficient modeling of long-tail noise. In this paper, we present Xuanwu VL-2B as a case study of how general multimodal models can be developed into an industrial-grade foundation model for content ecosystems. The model adopts a compact InternViT-300M + MLP + Qwen3 1.7B architecture, balancing fine-grained visual perception, language-semantic alignment, and deployment cost within an approximately 2B-parameter budget. To balance business specialization with the retention of general capabilities, we developed a data iteration and curation mechanism and trained the model through a progressive three-stage pipeline: pre-training, mid-training, and post-training. Ablation studies and offline business evaluations show that Xuanwu VL-2B achieves an average score of 67.90 across seven OpenCompass multimodal metrics (vs. 64.27 for InternVL 3.5 2B), an average recall of 94.38% over seven independent business moderation tasks, and a weighted overall recall of 82.82% on policy-violating text in challenging adversarial OCR scenarios, outperforming Gemini-2.5-Pro (76.72%). These results show that, under a limited parameter budget, Xuanwu VL-2B achieves a practical balance among business alignment, visual perception, general capability retention, and deployment cost.

3D Gaussian Splatting (3DGS) enables high-fidelity reconstruction of scene geometry and appearance. Building on this capability, inserting external mesh objects into reconstructed 3DGS scenes enables interactive editing and content augmentation for immersive applications such as AR/VR, virtual staging, and digital content creation. However, achieving physically consistent lighting and shadows for mesh insertion remains challenging, as it requires accurate scene illumination estimation and multi-view consistent rendering. To address this challenge, we present LightHarmony3D, a novel framework for illumination-consistent mesh insertion in 3DGS scenes. Central to our approach is our proposed generative module that predicts a full 360° HDR environment map at the insertion location via a single forward pass. By leveraging generative priors instead of iterative optimization, our method efficiently captures dominant scene illumination and enables physically grounded shading and shadows for inserted meshes while maintaining multi-view coherence. Furthermore, we introduce the first dedicated benchmark for mesh insertion in 3DGS, providing a standardized evaluation framework for assessing lighting consistency and photorealism. Extensive experiments across multiple real-world reconstruction datasets demonstrate that LightHarmony3D achieves state-of-the-art realism and multi-view consistency.

Routing is widely used to scale large language models, from Mixture-of-Experts gating to multi-model/tool selection. A common belief is that routing to a task ``expert'' activates sparser internal computation and thus yields more certain and stable outputs (the Sparsity--Certainty Hypothesis). We test this belief by injecting routing-style meta prompts as a textual proxy for routing signals in front of frozen instruction-tuned LLMs. We quantify (C1) internal density via activation sparsity, (C2) domain-keyword attention, and (C3) output stability via predictive entropy and semantic variation. On a RouterEval subset with three instruction-tuned models (Qwen3-8B, Llama-3.1-8B-Instruct, and Mistral-7B-Instruct-v0.2), meta prompts consistently densify early/middle-layer representations rather than increasing sparsity; natural-language expert instructions are often stronger than structured tags. Attention responses are heterogeneous: Qwen/Llama reduce keyword attention, while Mistral reinforces it. Finally, the densification--stability link is weak and appears only in Qwen, with near-zero correlations in Llama and Mistral. We present RIDE as a diagnostic probe for calibrating routing design and uncertainty estimation.

The AEC-Bench is a multimodal benchmark for evaluating agentic systems on real-world tasks in the Architecture, Engineering, and Construction (AEC) domain. The benchmark covers tasks requiring drawing understanding, cross-sheet reasoning, and construction project-level coordination. This report describes the benchmark motivation, dataset taxonomy, evaluation protocol, and baseline results across several domain-specific foundation model harnesses. We use AEC-Bench to identify consistent tools and harness design techniques that uniformly improve performance across foundation models in their own base harnesses, such as Claude Code and Codex. We openly release our benchmark dataset, agent harness, and evaluation code for full replicability at https://github.com/nomic-ai/aec-bench under an Apache 2 license.

Long-horizon dialogue systems suffer from semanticdrift and unstable memory retention across extended sessions. This paper presents a Multi-Layer Memory Framework that decomposes dialogue history into working, episodic, and semantic layers with adaptive retrieval gating and retention regularization. The architecture controls cross-session drift while maintaining bounded context growth and computational efficiency. Experiments on LOCOMO, LOCCO, and LoCoMo show improved performance, achieving 46.85 Success Rate, 0.618 overall F1 with 0.594 multi-hop F1, and 56.90% six-period retention while reducing false memory rate to 5.1% and context usage to 58.40%. Results confirm enhanced long-term retention and reasoning stability under constrained context budgets.

Large Language Models (LLMs) often experience performance degradation during long-running interactions due to increasing context length, memory saturation, and computational overhead. This paper presents an adaptive context compression framework that integrates importance-aware memory selection, coherence-sensitive filtering, and dynamic budget allocation to retain essential conversational information while controlling context growth. The approach is evaluated on LOCOMO, LOCCO, and LongBench benchmarks to assess answer quality, retrieval accuracy, coherence preservation, and efficiency. Experimental results demonstrate that the proposed method achieves consistent improvements in conversational stability and retrieval performance while reducing token usage and inference latency compared with existing memory and compression-based approaches. These findings indicate that adaptive context compression provides an effective balance between long-term memory preservation and computational efficiency in persistent LLM interactions

Prevailing 2D-centric visuomotor policies exhibit a pronounced deficiency in novel view generalization, as their reliance on static observations hinders consistent action mapping across unseen views. In response, we introduce GenSplat, a feed-forward 3D Gaussian Splatting framework that facilitates view-generalized policy learning through novel view rendering. GenSplat employs a permutation-equivariant architecture to reconstruct high-fidelity 3D scenes from sparse, uncalibrated inputs in a single forward pass. To ensure structural integrity, we design a 3D-prior distillation strategy that regularizes the 3DGS optimization, preventing the geometric collapse typical of purely photometric supervision. By rendering diverse synthetic views from these stable 3D representations, we systematically augment the observational manifold during training. This augmentation forces the policy to ground its decisions in underlying 3D structures, thereby ensuring robust execution under severe spatial perturbations where baselines severely degrade.

The aim of our paper is to render an object in 3-dimension using a set of its orthographic views. Corner detector (Harris Detector) is applied on the input views to obtain control points. These control points are projected perpendicular to respective views, in order to construct an envelope. A set of points describing the object in 3-dimension, are obtained from the intersection of these mutually perpendicular envelopes. These set of points are used to regenerate the surfaces of the object using computational geometry. At the end, the object in 3-dimension is rendered using OpenGL

This paper proposes the synthetic long-video meta-evaluation (SLVMEval), a benchmark for meta-evaluating text-to-video (T2V) evaluation systems. The proposed SLVMEval benchmark focuses on assessing these systems on videos of up to 10,486 s (approximately 3 h). The benchmark targets a fundamental requirement, namely, whether the systems can accurately assess video quality in settings that are easy for humans to assess. We adopt a pairwise comparison-based meta-evaluation framework. Building on dense video-captioning datasets, we synthetically degrade source videos to create controlled "high-quality versus low-quality" pairs across 10 distinct aspects. Then, we employ crowdsourcing to filter and retain only those pairs in which the degradation is clearly perceptible, thereby establishing an effective final testbed. Using this testbed, we assess the reliability of existing evaluation systems in ranking these pairs. Experimental results demonstrate that human evaluators can identify the better long video with 84.7%-96.8% accuracy, and in nine of the 10 aspects, the accuracy of these systems falls short of human assessment, revealing weaknesses in text-to-long-video evaluation.

Visual relocalization is a fundamental task in the field of 3D computer vision, estimating a camera's pose when it revisits a previously known scene. While point-based hierarchical relocalization methods have shown strong scalability and efficiency, they are often limited by sparse image observations and weak feature matching. In this work, we propose SplatHLoc, a novel hierarchical visual relocalization framework that uses Feature Gaussian Splatting as the scene representation. To address the sparsity of database images, we propose an adaptive viewpoint retrieval method that synthesizes virtual candidates with viewpoints more closely aligned with the query, thereby improving the accuracy of initial pose estimation. For feature matching, we observe that Gaussian-rendered features and those extracted directly from images exhibit different strengths across the two-stage matching process: the former performs better in the coarse stage, while the latter proves more effective in the fine stage. Therefore, we introduce a hybrid feature matching strategy, enabling more accurate and efficient pose estimation. Extensive experiments on both indoor and outdoor datasets show that SplatHLoc enhances the robustness of visual relocalization, setting a new state-of-the-art.

Adversarial robustness evaluation faces a critical challenge as new defense paradigms emerge that can exploit limitations in existing assessment methods. This paper reveals that Dummy Classes-based defenses, which introduce an additional "dummy" class as a safety sink for adversarial examples, achieve significantly overestimated robustness under conventional evaluation strategies like AutoAttack. The fundamental limitation stems from these attacks' singular focus on misleading the true class label, which aligns perfectly with the defense mechanism--successful attacks are simply captured by the dummy class. To address this gap, we propose Dummy-Aware Weighted Attack (DAWA), a novel evaluation method that simultaneously targets both the true label and dummy label with adaptive weighting during adversarial example synthesis. Extensive experiments demonstrate that DAWA effectively breaks this defense paradigm, reducing the measured robustness of a leading Dummy Classes-based defense from 58.61% to 29.52% on CIFAR-10 under l_infty perturbation (epsilon=8/255). Our work provides a more reliable benchmark for evaluating this emerging class of defenses and highlights the need for continuous evolution of robustness assessment methodologies.

Existing vision-and-language navigation (VLN) models primarily reason over past and current visual observations, while largely ignoring the future visual dynamics induced by actions. As a result, they often lack an effective understanding of the causal relationship between actions and how the visual world changes, limiting robust decision-making. Humans, in contrast, can imagine the near future by leveraging action-dynamics causality, which improves both environmental understanding and navigation choices. Inspired by this capability, we propose LatentPilot, a new paradigm that exploits future observations during training as a valuable data source to learn action-conditioned visual dynamics, while requiring no access to future frames at inference. Concretely, we propose a flywheel-style training mechanism that iteratively collects on-policy trajectories and retrains the model to better match the agent's behavior distribution, with an expert takeover triggered when the agent deviates excessively. LatentPilot further learns visual latent tokens without explicit supervision; these latent tokens attend globally in a continuous latent space and are carried across steps, serving as both the current output and the next input, thereby enabling the agent to dream ahead and reason about how actions will affect subsequent observations. Experiments on R2R-CE, RxR-CE, and R2R-PE benchmarks achieve new SOTA results, and real-robot tests across diverse environments demonstrate LatentPilot's superior understanding of environment-action dynamics in scene. Project page:https://abdd.top/latentpilot/

End-to-end multi-modal planning has been widely adopted to model the uncertainty of driving behavior, typically by scoring candidate trajectories and selecting the optimal one. Existing approaches generally fall into two categories: scoring a large static trajectory vocabulary, or scoring a small set of dynamically generated proposals. While static vocabularies often suffer from coarse discretization of the action space, dynamic proposals provide finer-grained precision and have shown stronger empirical performance on existing benchmarks. However, it remains unclear whether dynamic generation is fundamentally necessary, or whether static vocabularies can already achieve comparable performance when they are sufficiently dense to cover the action space. In this work, we start with a systematic scaling study of Hydra-MDP, a representative scoring-based method, revealing that performance consistently improves as trajectory anchors become denser, without exhibiting saturation before computational constraints are reached. Motivated by this observation, we propose SparseDriveV2 to push the performance boundary of scoring-based planning through two complementary innovations: (1) a scalable vocabulary representation with a factorized structure that decomposes trajectories into geometric paths and velocity profiles, enabling combinatorial coverage of the action space, and (2) a scalable scoring strategy with coarse factorized scoring over paths and velocity profiles followed by fine-grained scoring on a small set of composed trajectories. By combining these two techniques, SparseDriveV2 achieves 92.0 PDMS and 90.1 EPDMS on NAVSIM, with 89.15 Driving Score and 70.00 Success Rate on Bench2Drive with a lightweight ResNet-34 as backbone. Code and model are released at https://github.com/swc-17/SparseDriveV2.