Daily ArXiv / October 07, 2026

Personalized paper radar

A focused reading queue selected from today's ArXiv feed, ranked by topic fit, novelty, and configured author matches.

Relevant papers 15

Top score 15

Average score 11.3

Source ArXiv

Abstract word clouds

Today

agentattentionchartchunkcolorcommcommunicationcomplementaryconsistencyconstructioncontentcontrolcostdiffusiondistributiondomaindynamicevidencefixedgenerationinteractionlatentmotionmultimodalphysicalpromptreasoningreferencereliablesensorshiftsimulationsolversparsespatialsteptargettargetedtextbftokenunderstandingverificationvideovisualworld

Past month

actionagentannotationcamerachallengingchangeconsistencycontentcontrolcostdensedetectiondomaindynamicenvironmentevidencefoundationgenerationgeometricgeometryinferenceinteractionlanguagelatentmemorymotionmultimodalmultipleobjectobservationperceptionphysicalpipelinepointpolicyproducequeryquestionreasoningreconstructionreferenceregionsamescenesemanticsourcespacespatialsupervisionsupporttargettemporaltokentrajectoryunderstandingunifiedvideovision-languagevisualworld

Reading Queue

cs.CV

11 papers

Embodied AI

1 R2RI: A Multi-View Event and RGB Dataset for Robot-to-Robot Interaction Gabriele Magrini, Riccardo Catalini, Federico Becattini, Guido Borghi, Pietro Pala, Roberto Vezzani, Lorenzo Seidenari

Embodied AI Robotics Benchmark Egocentric Vision

cs.CV cs.RO

Paper 1 / arXiv:2610.07117 Open arXiv

Relevance 8 Novelty 7

Why selected: Strong match to criterion 3: a new embodied-AI benchmark/dataset for robot-to-robot interaction with egocentric and exocentric views, including event and RGB modalities.

Understanding and modeling interactions between autonomous agents is a fundamental challenge in robotics, with broad implications for collaborative systems, social robotics, and human-robot coexistence. Although the study of robot interactions has emerged as a compelling research direction, progress has been severely hampered by the absence of large-scale benchmarks. In this paper, we introduce Robot-to-Robot Interaction (R2RI), the first dataset specifically designed to address the Robot-Robot Interaction (RRI) task. R2RI consists of different humanoid robots and realistic interactions modeled on real human social behaviors. Complementary viewpoints are available, \textit{i.e.}, an egocentric perspective from each robot's onboard sensors, and an exocentric perspective from external fixed cameras, thus enabling rich spatial and contextual understanding of the interaction dynamics. The dataset comprises more than $6.5$M frames and $\approx5000$ videos at $120$ fps, including Event and RGB domains. We investigate pros and cons of each domain, comparing state-of-the-art approaches for a number of key sensing and interaction based tasks. We publicly release the dataset and its annotations for all tasks and modalities at https://github.com/MagriniGabriele/R2RI.

Embodied World Models

3 Artemis: Geometry-Grounded Multi-Agent Driving World Models with Shared 3D State and Progressive Memory Update Sitian Shen, Jiuming Liu, Mengmeng Liu, Yian Wang, Michael Ying Yang, Francesco Nex, Hao Cheng, Daniele De Martini, Ayush Tewari, Per Ola Kristensson

Embodied World Models Driving Simulation 3D Geometry

cs.CV

Paper 3 / arXiv:2610.07031 Open arXiv

Relevance 6 Novelty 7

Why selected: Matches criterion 3: a new embodied driving world-model method with explicit 3D state, multi-agent geometry grounding, and a new CARLA-based dataset.

Recent video world models have witnessed the paradigm shift from single-agent to multi-agent involvements, which can reveal more complicated dynamics and cross-agent interaction in the real world. However, existing approaches commonly adopt implicit inter-agent communications via cross attention, which lack explicit geometry constraints and unified 3D state, thereby leading to poor multi-view consistency and struggling with recovering out-of-sight agents. In addition, most of them assume a static background, failing to represent uncontrolled background dynamics. To address these problems, we propose Artemis: a geometry-grounded multi-agent world model with explicit memory sharing. An explicit 3D world map is reconstructed from multi-agent observations to enforce a unified 3D state across agents, offering high cross-view consistency. Specifically, an action-guided geometric injection module is developed to simultaneously render decomposed foreground-background control maps, which are then injected into a diffusion transformer through a designed GeoAdapter block. Compared to previous methods assuming static-only background, our GeoAdapter can also distinguish uncontrolled non-agent dynamics, which are conditioned on their own multi-frame history positions to provide consistent motion cues. Keyframes selected from progressive video rollouts are used to progressively update the reconstructed 3D world maps. To effectively capture complex dynamic patterns, we curate a novel dataset sampled from the CARLA simulator called MA-CARLA. Extensive experiments demonstrate the superiority of our proposed method in terms of visual fidelity and cross-view consistency in the generated videos. In addition, our Artemis can support simultaneous multi-modal rollouts with both 2D video and 3D point map maintenance, scale to scenarios beyond two agents and multi-camera setting.

Neural Simulation

5 PhysLDM: Latent Diffusion for High-Fidelity Deformable Simulation Yu Zhang, Xudong Xu, Xingang Pan

Neural Simulation Latent Diffusion Deformable Physics

cs.CV cs.GR

Paper 5 / arXiv:2610.07609 Open arXiv

Relevance 4 Novelty 8

Why selected: Partial match to criterion 3 via physical simulation, and to criterion 4 only loosely through generative modeling; it is mainly a neural simulation / latent-diffusion paper rather than vision foundation models.

Neural simulation of high-fidelity deformable bodies is a foundational challenge in computer graphics and physical AI. Long-horizon prediction for high-resolution 3D volumetric meshes is hard: autoregressive methods are susceptible to error accumulation, while direct multi-frame prediction at native resolution is computationally prohibitive. This motivates a compact spatiotemporal latent representation, which is largely unexplored for mesh-based volumetric physics. Meanwhile, it remains unclear whether deterministic regression or generative diffusion is the more appropriate predictive paradigm. To address these coupled challenges, we introduce PhysLDM, a unified latent-diffusion paradigm for one-shot volumetric deformable simulation. Its core is a holistic spatiotemporal VAE that avoids the "staircase" artifacts of standard temporal compression (as in common video VAEs), achieving ~2.48 mm reconstruction precision on meter-scale scenes at up to 78x token compression. Based on this reliable latent space, we systematically compare regression and diffusion methods. Our experiments uncover a key modeling insight: complex deformable dynamics are often chaotic, and in this regime deterministic regression tends to produce non-physical averages, whereas diffusion better models their distribution. Accordingly, we employ a latent diffusion model that effectively learns from the chaotic data to generate physically plausible trajectories. Trained purely kinematically on an Objaverse-scale dataset, a single PhysLDM generalizes zero-shot to unseen OOD datasets (GSO and Toys4K). Its differentiability further enables efficient solution of inverse problems and higher-order design optimization. To our knowledge, PhysLDM is the first high-fidelity spatiotemporal autoencoder and latent-diffusion paradigm for volumetric deformable dynamics, offering a scalable and robust approach to neural simulation.

Spatial Understanding

6 Structure-aware Keypoint Localization for Videofluoroscopic Swallowing Study Kai Zhou, Chuanshen Chen, Runhao Zeng, Meng Dai, Yifan Yang, Jinwu Hu, Daiyuan Li, Mingkui Tan, Fei Liu

Spatial Understanding Keypoint Localization Semi-Supervised Learning

cs.CV

Paper 6 / arXiv:2610.07726 Open arXiv

Relevance 5 Novelty 6

Why selected: Close match to criterion 1: it proposes a structure-aware method for spatial keypoint localization, though in a medical X-ray setting rather than embodied agents.

Videofluoroscopic Swallowing Study (VFSS) is one of the gold standard for diagnosing swallowing disorders, providing dynamic X-ray imaging of the swallowing process. Automated kinematic analysis in VFSS relies fundamentally on precise anatomical keypoint localization. However, existing studies focus on limited keypoints (e.g., cervical vertebrae or the hyoid) and overlook critical regions such as the soft palate, while annotating only active swallowing segments and ignoring abundant non-swallowing data, resulting in poor data efficiency. Moreover, leveraging this unlabeled data via standard semi-supervised learning is suboptimal, as generic methods are prone to spatial bias. In medical X-rays with fixed layouts, models tend to memorize absolute coordinates rather than understanding anatomical structures. To tackle these challenges, we introduce VFSSKep, a novel dataset that extends annotations to the soft palate and incorporates large-scale unlabeled data. We further propose S$^3$KL, a Structure-aware Semi-Supervised Keypoint Localization framework designed to overcome spatial bias. It integrates a Structure-Aware Learning strategy to extract high-resolution structural cues for structure-aware representation learning, and a Structural Representation Consistency Learning strategy with block shuffling to enforce invariant structural recognition. Experiments show our method achieves state-of-the-art semi-supervised performance, even with unlabeled and 25% labeled data surpassing fully supervised learning with 100% labeled data. Code and data will be made publicly available at: https://github.com/kaai520/S3KL.

Vision Foundation Models

8 On Color Alignment in VAE Latent Spaces and Its Applications Julian D. Santamaria, Kai Wang, Jes\'us Malo, Javier Vazquez-Corral, Alexandra G\'omez-Villa

Vision Foundation Models VAE Latent Space Image Editing & Control

cs.CV cs.LG

Paper 8 / arXiv:2610.07072 Open arXiv

Relevance 5 Novelty 6

Why selected: Matches criterion 4 very closely: analyzes a latent-space property of VAEs used in text-to-image foundation models and turns it into practical control applications.

Variational autoencoders (VAEs) are a key part of modern text-to-image models, which generate images within their latent space. VAEs are known to disentangle the main factors of variation in the data, and color is known to be one of the most structured of these in natural images: decorrelating it yields one luminance axis and two opponent-color axes. Color should therefore be expected to emerge as a distinct factor in the VAE latent space. Yet how these latent spaces represent color remains largely unexplored. In this work, we show that the VAEs of text-to-image models share a color subspace aligned with brightness and opponent-colors. Through a linear approximation of the encoder and targeted latent steering, we find this subspace consistently across a broad range of VAEs, from SD1.5 to FLUX.2 and Z-Image. Building on this characterization, we propose three applications: ColorTuning, which achieves state-of-the-art in precise numerical color generation on the fine-grained CSS3/X11 system of GenColorBench, saturation control, to adjust the global chromatic intensity, and color transfer, to change the palette to match a reference. The code and models are publicly available at https://julian075.github.io/Color_Subspace/

Video World Models

9 World Models' Last Exam in Physics Mingju Gao, Qingle Liu, Yuzhao Peng, Xinjie Lin, Ziming Qin, Zheng Jiang, Wenyi Li, Calvin Xiao, Youjie Zheng, Kaisen Yang, Qinhuai Na

Video World Models Physics Consistency Benchmark & Evaluation

cs.CV

Paper 9 / arXiv:2610.08791 Open arXiv

Relevance 5 Novelty 6

Why selected: Matches criterion 3: a new benchmark for evaluating physical consistency in video world models, focused on measurable physical properties rather than reference-video similarity.

Video world models can produce visually convincing yet physically inconsistent sequences, raising concerns about their reliability for prediction and planning in embodied AI systems. Existing evaluations often rely on model-based judgments or reference videos, while direct physical tests largely focus on mechanics. We introduce World Models' Last Exam in Physics, a measurement-based benchmark for evaluating physical consistency in video world models. The benchmark comprises 40 controlled tasks spanning mechanics, optics, fluids, thermal and phase-change phenomena, electromagnetism, and surface tension. Each task pairs an initial image and a generation prompt with predefined physical criteria, enabling interpretable tests of observable physical relationships without requiring reference videos. Its evaluator combines task-observability screening with task-specific quantitative physical measurements. Experiments on eight video generation models across 1,280 videos reveal persistent physical inconsistencies and substantial variation across tasks, with the best model achieving an overall score of 57.76 out of 100. Evaluation on synthetic videos with known physical relationships provides evidence for the validity of the measurement module under controlled conditions. The evaluator also achieves higher agreement with human judgments than a direct vision-language model baseline in both within-task rankings and pairwise comparisons. By combining coverage across physical domains with scores grounded in measurable evidence and explicit measurement limitations, the benchmark provides an interpretable basis for diagnosing physical inconsistencies and tracking progress toward physically consistent video world models.

Interactive Video World Models

11 CtrlCache: Accelerating Interactive Video World Models with Control-Aware Caching Shangye Song, Dong Gong, Hong Jia, Yun Sing Koh, Xinyu Zhang

Interactive Video World Models Efficiency & Caching Embodied Simulation

cs.CV

Paper 11 / arXiv:2610.08777 Open arXiv

Relevance 5 Novelty 6

Why selected: Matches criterion 3 very closely: a new method for interactive video world models with control-aware caching, explicitly aimed at efficient embodied/simulator-like rollout generation.

Interactive video world models need to generate each video chunk efficiently while responding faithfully to user controls. Many systems use chunk-wise autoregressive generation with few-step denoising, but each chunk still requires several costly denoising iterations. Training-free caching can reduce this cost, yet existing policies make reuse decisions primarily from model-internal denoising dynamics and do not explicitly account for control transitions. Actually, interactive generation explicitly exposes a signal they do not use: the controls for a chunk arrive before it is denoised, so a schedule derived from them costs no forward pass. To this end, we analyze adjacent chunks under different control regimes and find that structural similarity drops around action changes, while low-frequency structure remains more persistent than high-frequency detail. Motivated by these observations, we propose CtrlCache, a training-free control-aware caching framework that adapts computation to the current control sequence. Specifically, the action-aware scheduling and refresh policy detects action changes across and within chunks, and labels each chunk as initial, transition, turning, or steady state. At one selected interior denoising step, initial and transition chunks retain full computation, while turning and steady chunks reuse the transformer residual from the most recent fully computed step in the same chunk. To exploit the persistence of low-frequency structure during steady interaction, we further introduce a frequency-mixed history prior guidance that incorporates complementary information from the preceding clean latent without an additional DiT forward pass. Evaluated on Matrix-Game 2.0 and LingBot-World v1/v2, CtrlCache achieves 1.21x to 1.41x DiT-backbone speedups without model retraining while improving WBench Overall scores over original inference across all three models.

Image Generation

12 RefRoute: Decoupling Conditioning Cost from References via Compact Residual Conditioning and Spatial Routing Wanning He, Yuyao Zhang, Yu-Wing Tai

Image Generation Reference Conditioning Benchmark & Evaluation

cs.CV cs.LG

Paper 12 / arXiv:2610.07720 Open arXiv

Relevance 5 Novelty 5

Why selected: Matches criterion 4: it is a diffusion/vision-model application for scalable multi-reference image generation, with a new benchmark and routing-based method.

Multi-reference image generation requires preserving the appearance of multiple subjects while composing them into a coherent scene. However, existing diffusion transformers commonly encode references as dense visual token grids and jointly process them with global attention, making conditioning increasingly expensive as the number and resolution of references grow. We present RefRoute, a framework that addresses both reference representation cost and attention overhead through two complementary mechanisms. Compact residual conditioning combines low-resolution latent tokens with lightweight residual features extracted from full-resolution pixels, reducing reference token counts while retaining fine-grained appearance cues. Condition routing and attention routing align reference tokens with their assigned target regions and restrict cross-reference interactions, while allowing selective reference access beyond region boundaries for scene integration. We further introduce RefRoute-Data for training many-reference generation models and ManyRef100, a benchmark spanning human, object, and mixed compositions with 10-17 references. After many-reference fine-tuning, RefRoute achieves an overall Weighted-Ref-VIEScore of 36.06 on ManyRef100, compared with 8.88 for FLUX.2-Klein-9B. Separate inference-cost evaluations show substantially slower latency growth as the reference count increases: at 16 references, our 50-step and 4-step configurations achieve $18.3\times$ and $14.2\times$ speedups over their corresponding FLUX baselines, respectively. These results establish compact reference representations and spatially routed attention as an effective approach to scalable many-reference image generation.

Video Generation

13 Diverse Motion Customization via Control-based Dynamic Optimization Youngyoon Choi, Kihyun Kim, Jeongwoo Shin, Joonseok Lee

Video Generation Motion Customization Generative Modeling

cs.CV cs.AI

Paper 13 / arXiv:2610.07911 Open arXiv

Relevance 3 Novelty 7

Why selected: Partial match to criterion 4 through video generative modeling, but it is not a vision foundation model paper and does not target embodied AI or spatial intelligence.

Despite recent advances in video generation, motion customization remains challenging due to content leakage, where appearance attributes from the reference video unintentionally propagate into the generated output. We identify this issue as a consequence of the generative process collapsing toward the reference video, which arises from formulating the learning objective as a direct regression on the reference. To address this, we propose Control-based Motion Customization (CMC), a principled training framework that is structurally robust to content leakage. Our key idea is to steer generative dynamics toward desired motion while avoiding collapse toward the reference video, which we formalize using Stochastic Optimal Control (SOC). Under this formulation, customized videos acquire the target motion yet remain within the pre-trained model's prompt-conditional distribution, where appearance is determined by the text prompt rather than the reference video. Furthermore, to improve efficiency, we tailor the SOC formulation to motion customization by eliminating the need for an explicit reward and introducing a timestep-adaptive motion cost that focuses only on early generative stages, accelerating training by 2.5 times. Extensive experiments demonstrate that CMC effectively mitigates content leakage and achieves competitive motion fidelity while preserving the diversity of the base model across diverse scenarios.

Cooperative Perception

14 Sparse2comm: Towards Robust Cooperative 3D Object Detection Lei Yang, Boqi Li, Chunmian Lin, Li Wang, Ziying Song, Shaoqing Xu, Heye Huang, Haibao Yu, Chen Lv

Cooperative Perception 3D Object Detection Robust Sensor Fusion

cs.CV

Paper 14 / arXiv:2610.08573 Open arXiv

Relevance 5 Novelty 5

Why selected: Matches criterion 3: a new cooperative 3D perception method for autonomous driving, with robustness to communication degradation and spatial misalignment.

Cooperative perception improves autonomous driving by sharing complementary observations among vehicles and roadside infrastructure for 3D object detection. However, practical deployment is constrained by limited bandwidth and unreliable cooperation, where packet loss, transmission delay, and spatial misalignment jointly degrade the cooperative feature stream. Existing methods often reduce communication cost or compensate for one degradation type, leaving coupled disturbances insufficiently addressed. To address this problem, we propose Sparse2comm, a bandwidth-efficient and robust cooperative 3D object detection framework that treats unreliable cooperation as progressive restoration over degraded cooperative features. Sparse Feature Encoding first encodes communication as randomly mask-sampled foreground features transmitted by collaborating agents, from which the ego vehicle reconstructs dense semantic representations. This sparse-to-dense mechanism learns to infer missing object-centric content from sparse observations, enabling ultra-low-bandwidth communication and packet-loss recovery within the same representation. On the semantically restored features, Latency-Aware Alignment predicts motion flow to compensate delayed messages, and Self-Calibrating Fusion estimates residual spatial offsets in a self-supervised manner before adaptive cross-agent fusion. Sparse2comm therefore restores semantic completeness, temporal consistency, and spatial alignment in an ordered pipeline. Extensive experiments on DAIR-V2X, OpenV2V, and V2V4Real show that Sparse2comm maintains competitive clean accuracy and consistently improves robustness under individual and mixed real-world degradations. Compared with the selective feature communication baseline Where2comm, Sparse2comm improves mixed-setting AP@0.5/AP@0.7 by +20.15/+11.79, +12.66/+11.07, and +15.36/+12.61 on the three datasets, respectively.

Test-Time Adaptation

15 Test-Time Adaptation of Quantized ViTs via Single-Pass Quantizer-Aligned Recalibration Hyeongheon Cha, Young D. Kwon, Sung-Ju Lee

Test-Time Adaptation Model Quantization Vision Transformers

cs.CV cs.AI

Paper 15 / arXiv:2610.08358 Open arXiv

Relevance 3 Novelty 4

Why selected: Partial fit to criterion 4 only in the sense that it targets quantized ViTs, but it is primarily a test-time adaptation method rather than a vision foundation model application paper.

Post-training quantization is a standard route to fitting vision transformers (ViTs) into edge compute and memory budgets, yet quantized models become especially brittle under distribution shift. Test-time adaptation (TTA) addresses such shifts without labels, but most existing approaches are poorly aligned with the constraints of quantized inference. Prevailing TTA methods recover accuracy through backpropagation, while backprop-free methods often still incur overhead from extra forward passes or parameter updates, and lightweight feature- or logit-level methods recover only part of the loss. Across these approaches, a quantization-specific failure mode that amplifies the drop is not directly targeted: under shift, activations occupy frozen quantizers' calibrated ranges differently, distorting their code distribution. We propose Quantizer-Aligned Recalibration (QuAR), a single-pass TTA method tailored to quantized ViTs that neither backpropagates nor updates any model parameters. QuAR recalibrates activations at the input to a frozen quantizer, mapping the test stream's running per-channel statistics back toward the source calibration. On ImageNet-C with ViT-B, QuAR achieves the highest mean accuracy among state-of-the-art backprop-free TTA methods at 3-, 4-, 6- and 8-bit weight/activation precision, outperforming the strongest baseline by 2.28 points at 8 bits and 4.00 at 3 bits, with 46% lower latency and a memory overhead of only 0.17 MB (0.01% of peak inference memory). Analysis and diagnostics trace the gain to a reduced per-channel mismatch at these quantizers, which restores the code distribution the baselines leave unchanged or distort further. A single fixed configuration remains ahead across continual streams, non-i.i.d. label shift, seven out-of-distribution suites, and three other backbones.

cs.AI

4 papers

MLLMs

2 ChartBmkAgent: Harness-Governed Multi-Agent Construction of Chart QA Benchmarks from Sparse Error-Taxonomy Specifications Langxi Huang, Pingping Zhang, Lanyun Zhu, Chunyang Jiang, Jiawei Shao, Haocheng Yuan, Peilin Chen

MLLMs Benchmark Construction Chart QA

cs.AI

Paper 2 / arXiv:2610.08106 Open arXiv

Relevance 7 Novelty 7

Why selected: Strong match to criterion 2: it targets MLLMs and builds a new benchmark-construction pipeline for chart QA diagnostics.

Multimodal large language models (MLLMs) advance rapidly, while conventional benchmark development lags behind, delaying investigation of newly observed capability gaps. Such investigation requires an expressive task format and an on-demand construction process: information-rich charts make chart question answering (Chart QA) suitable for probing coupled perception and reasoning. Automated Chart QA construction is intended to shorten the benchmark-development cycle by turning identified gaps into targeted samples on demand. Current methods, however, commonly separate target guidance from scratch generation: target-guided systems often require prepared data, charts, or templates, while scratch-generation systems primarily ensure artifact validity, without explicitly controlling whether newly synthesized requirements and content remain aligned with an externally specified diagnostic target. We introduce ChartBmkAgent, which turns an identified capability gap into targeted diagnostic evidence by constructing complete Chart QA samples from sparse error-taxonomy specifications. Throughout construction, a central harness governs specialized agents, requires stage-specific evidence of alignment with the original error category, and records the basis for each acceptance decision. On 300 taxonomy-wide samples, MLLM accuracies ranged from 32.7% to 84.3% with distinct category profiles, showing that generated samples reveal capability differences. Across three source-model comparisons, targeted follow-ups scored 50.0% versus 82.2% on matched controls ($p=8.96\times10^{-6}$); all six cross-model comparisons had the same direction, demonstrating targeted validation and diagnostic-data generation. Multiple evaluator models assessed whether each sample tested its specified error category; 86.4% met this criterion, providing empirical evidence of target preservation.

Embodied AI Benchmark

4 WorldSolver: Can LLM Agents Simulate the Physical Dynamics via Solver Generation? Siru Jiang, Yongzhe Lyu, Shuo Lu, Yubin Wang, Yuxiang Zhang, Yue Liao, Bin Wang, Jian Liang, Tieniu Tan

Embodied AI Benchmark Physics Simulation Agentic Code Generation

cs.AI

Paper 4 / arXiv:2610.08720 Open arXiv

Relevance 6 Novelty 7

Why selected: Matches criterion 3 very closely: a new embodied/simulation-oriented benchmark for solver generation in physical dynamics, targeting a novel angle largely ignored by prior work.

LLM-based agents are increasingly advancing scientific and engineering problem solving, with physics simulation emerging as a challenging yet practical testbed for reproducing complex physical phenomena with application in embodied AI, games and films. As the workhorse of such simulation, a solver computes how the state of a dynamic system evolves over time. Building such solvers requires physical understanding to identify appropriate models, mathematical reasoning to formulate the underlying dynamics, and software engineering to implement them as executable code, yet this capability of LLM agents remains underexplored. To this end, we introduce WorldSolver, a benchmark of 168 simulation tasks derived from physical phenomena in 61 classic computer graphics papers, spanning 7 physical domains. Each task contains a code scaffold that provides a fixed simulation environment for the scene, with the solver implementation left for the agent to complete. Specifically, we evaluate them along three dimensions: Execution Checks for successful execution, Visual Fidelity for reproducing the intended dynamic behavior in the rendered simulation, and Physical Plausibility for physics-grounded verification of the generated dynamics. Experiments on frontier agents reveal that producing executable solvers is difficult itself, and satisfying visual and physical correctness is even harder. GPT-5.6-Sol and Claude-Opus-5 perform comparatively better than the other evaluated agents, yet achieve overall scores of only 48.7% and 46.7%, respectively. WorldSolver is an early step toward agentic solver generation, and we hope it helps drive progress toward agents that can faithfully simulate the dynamic physical world. Code is available at https://github.com/sirujiang/WorldSolver.

Vision-Language Models

7 When to Rethink: Learning Multi-Perspective Self-Verification for Vision-Language Models Ziquan Zhu, Hanruo Zhu, Si-Yuan Lu, Morris Yu-Chao Huang, Yicheng Lin, Wei Han, Tianlong Chen, Mingyuan Wu, Hanchao Yu, Gaojie Jin, Lu Liu, Bo Sun, Tianjin Huang

Vision-Language Models Self-Verification Multimodal Reasoning

cs.AI

Paper 7 / arXiv:2610.07018 Open arXiv

Relevance 5 Novelty 6

Why selected: Matches criterion 2: a new vision-language model self-verification framework with multi-perspective rethinking for more reliable multimodal reasoning.

Vision-language models (VLMs) have achieved strong performance in multimodal reasoning, yet they remain prone to generating plausible but incorrect answers. Self-verification offers a practical way to improve answer reliability without relying on external judges, but existing methods typically depend on a single verification criterion or fixed prompt, resulting in incomplete and unstable reliability estimates. We first systematically analyze how verifier capability and prompt design affect verification performance. Our findings show that stronger verifiers provide more reliable judgments, while verification performance is highly sensitive to prompt choice, with no single prompt consistently dominating across tasks. Guided by these findings, we propose \texttt{MOTIVE}, a \textbf{M}ulti-View Self-Verificati\textbf{O}n wi\textbf{T}h Rel\textbf{I}ability-Guided Selecti\textbf{VE} Rethinking framework for reliable multimodal reasoning. \texttt{MOTIVE} evaluates each candidate answer from complementary verification perspectives and learns a correctness-aligned reliability score through correctness-grounded multi-view verification learning. During inference, this score governs an accept-or-rethink decision, allowing reliable answers to be returned directly while uncertain ones trigger history-guided rethinking. Extensive experiments across diverse multimodal benchmarks and VLM backbones demonstrate that \texttt{MOTIVE} consistently outperforms strong self-verification and self-correction baselines. Further results show that reliable verification improves accept-or-rethink decisions and reduces unnecessary reasoning turns, enabling more reliable and efficient self-verification without an external judge.

Benchmark & Evaluation

10 How Well Do LLMs Reason with Noisy Evidence? An Active Visual Reasoning Benchmark Bach Nguyen, Zhaonan Li, Mau Son Nguyen, Sanika Chavan, Nilay Kumar, Hong Anh Nguyen, Khoa Vo, Ben Zhou

Benchmark & Evaluation Active Visual Reasoning Noisy VLMs

cs.AI

Paper 10 / arXiv:2610.07751 Open arXiv

Relevance 5 Novelty 6

Why selected: Matches criterion 3: introduces a new benchmark for active visual reasoning under noisy evidence, with a simulator-like sensor setup for studying uncertainty-aware reasoning.

Real-world reasoning rarely reduces to static question answering: agents must actively gather information from tools and sensors that are often noisy and unreliable. Yet most existing active reasoning benchmarks assume that environmental feedback is trustworthy, or introduce noise without exposing an explicit, calibrated uncertainty signal, leaving open how LLMs should reason when the evidence itself is uncertain. We introduce VisualNoiseQA, a novel benchmark for active reasoning under noisy visual feedback. A text-only LLM must solve VQA problems by iteratively querying a fixed, off-the-shelf VLM treated as a stochastic visual sensor. For each query, we draw multiple samples and expose an empirical uncertainty signal via self-consistency, enabling the reasoner to probe from different angles and decide what to ask next and when to stop. Our construction is automatic and scalable: starting from diverse VQA sources and two noisy VLMs, we retain only questions where the sensor is inconsistent yet human-solvable. We evaluate multiple LLM reasoners on 1,000 instances spanning perception, chart understanding, and knowledge-intensive reasoning. VisualNoiseQA thus provides a controlled playground to study how different LLMs exploit uncertainty signals for robust reasoning.

Past ArXiv

October 05, 2026 October 02, 2026 October 01, 2026 September 30, 2026 September 29, 2026 September 28, 2026 September 25, 2026 September 24, 2026 September 23, 2026 September 22, 2026 September 21, 2026 September 18, 2026 September 17, 2026 September 16, 2026 September 15, 2026 September 14, 2026 September 12, 2026 September 11, 2026 September 10, 2026 September 07, 2026

Paper selection prompt

 1. New methodological improvements to spatial understanding, spatial intelligence on embodied agents;
 2. Shows new VLLMs (visual large language models) or MLLMs (multi-modal large language models)
 3. Embodied AI papers on buliding new benchmark (simulator related) or new methods. These papers should focus on novel angles that previous work ignored.
 4. Vision foundation models related and its applications.

 In suggesting papers to your friend, remember that he enjoys papers on computer vision and machine learning, and generative modeling in multi-modal learning.
 Your friend also likes learning about surprising empirical or insightful results in vision-language models or embodied AI, as well as clever statistical tricks.