While recent multimodal large language models (MLLMs) have made impressive strides, they predominantly employ a conventional autoregressive architecture as their backbone, leaving significant room to explore effective and efficient alternatives in architectural design. Concurrently, recent studies have successfully applied discrete diffusion models to various domains, such as visual understanding and image generation, revealing their considerable potential as a promising backbone for multimodal systems. Drawing inspiration from these pioneering research, we introduce Omni-Diffusion, the first any-to-any multimodal language model built entirely on mask-based discrete diffusion models, which unifies understanding and generation across text, speech, and images. Omni-Diffusion employs a unified mask-based discrete diffusion model to directly capture the joint distribution over discrete multimodal tokens. This approach supports not only bimodal tasks but also more complex scenarios involving multiple modalities. On a diverse set of benchmarks, our method outperforms or performs on par with existing multimodal systems that process two or more modalities, highlighting the significant promise of diffusion models in powering the next generation of multimodal foundation models. Project webpage: https://omni-diffusion.github.io.

The integration of Large Language Models (LLMs) into autonomous driving has attracted growing interest for their strong reasoning and semantic understanding abilities, which are essential for handling complex decision-making and long-tail scenarios. However, existing methods typically feed LLMs with tokens from multi-view and multi-frame images independently, leading to redundant computation and limited spatial consistency. This separation in visual processing hinders accurate 3D spatial reasoning and fails to maintain geometric coherence across views. On the other hand, Bird's-Eye View (BEV) representations learned from geometrically annotated tasks (e.g., object detection) provide spatial structure but lack the semantic richness of foundation vision encoders. To bridge this gap, we propose BEVLM, a framework that connects a spatially consistent and semantically distilled BEV representation with LLMs. Through extensive experiments, we show that BEVLM enables LLMs to reason more effectively in cross-view driving scenes, improving accuracy by 46%, by leveraging BEV features as unified inputs. Furthermore, by distilling semantic knowledge from LLMs into BEV representations, BEVLM significantly improves closed-loop end-to-end driving performance by 29% in safety-critical scenarios.

Obstacle avoidance in unmanned aerial vehicles (UAVs), as a fundamental capability, has gained increasing attention with the growing focus on spatial intelligence. However, current obstacle-avoidance methods mainly depend on limited field-of-view sensors and are ill-suited for UAV scenarios which require full-spatial awareness when the movement direction differs from the UAV's heading. This limitation motivates us to explore omnidirectional obstacle avoidance for panoramic drones with full-view perception. We first study an under explored problem setting in which a UAV must generate collision-free motion in environments with obstacles from arbitrary directions, and then construct a benchmark that consists of three representative flight tasks. Based on such settings, we propose Fly360, a two-stage perception-decision pipeline with a fixed random-yaw training strategy. At the perception stage, panoramic RGB observations are input and converted into depth maps as a robust intermediate representation. For the policy network, it is lightweight and used to output body-frame velocity commands from depth inputs. Extensive simulation and real-world experiments demonstrate that Fly360 achieves stable omnidirectional obstacle avoidance and outperforms forward-view baselines across all tasks. Our model is available at https://zxkai.github.io/fly360/

Surgeons don't just see -- they interpret. When an expert observes a surgical scene, they understand not only what instrument is being used, but why it was chosen, what risk it poses, and what comes next. Current surgical AI cannot answer such questions, largely because training data that explicitly encodes surgical reasoning is immensely difficult to annotate at scale. Yet surgical video lectures already contain exactly this -- explanations of intent, rationale, and anticipation, narrated by experts for the purpose of teaching. Though inherently noisy and unstructured, these narrations encode the reasoning that surgical AI currently lacks. We introduce SUREON, a large-scale video QA dataset that systematically harvests this training signal from surgical academic videos. SUREON defines 12 question categories covering safety assessment, decision rationale, and forecasting, and uses a multi-agent pipeline to extract and structure supervision at scale. Across 134.7K clips and 170 procedure types, SUREON yields 206.8k QA pairs and an expert-validated benchmark of 354 examples. To evaluate the extent to which this supervision translates to surgical reasoning ability, we introduce two models: SureonVLM, a vision-language model adapted through supervised fine-tuning, and SureonVLM-R1, a reasoning model trained with Group Relative Policy Optimization. Both models can answer complex questions about surgery and substantially outperform larger general-domain models, exceeding 84% accuracy on the SUREON benchmark while outperforming general-domain models on standard surgical perception tasks. Qualitative analysis of SureonVLM-R1 reveals explicit reasoning behavior, such as inferring operative intent from visual context.

Machine-learning interatomic potentials (MLIPs) have advanced rapidly, with many top models relying on strong physics-based inductive biases. However, as models scale to larger systems like biomolecules and electrolytes, they struggle to accurately capture long-range (LR) interactions, leading current approaches to rely on explicit physics-based terms or components. In this work, we propose AllScAIP, a straightforward, attention-based, and energy-conserving MLIP model that scales to O(100 million) training samples. It addresses the long-range challenge using an all-to-all node attention component that is data-driven. Extensive ablations reveal that in low-data/small-model regimes, inductive biases improve sample efficiency. However, as data and model size scale, these benefits diminish or even reverse, while all-to-all attention remains critical for capturing LR interactions. Our model achieves state-of-the-art energy/force accuracy on molecular systems, as well as a number of physics-based evaluations (OMol25), while being competitive on materials (OMat24) and catalysts (OC20). Furthermore, it enables stable, long-timescale MD simulations that accurately recover experimental observables, including density and heat of vaporization predictions.

Deep reinforcement learning agents are often misaligned, as they over-exploit early reward signals. Recently, several symbolic approaches have addressed these challenges by encoding sparse objectives along with aligned plans. However, purely symbolic architectures are complex to scale and difficult to apply to continuous settings. Hence, we propose a hybrid approach, inspired by humans' ability to acquire new skills. We use a two-stage framework that injects symbolic structure into neural-based reinforcement learning agents without sacrificing the expressivity of deep policies. Our method, called Hybrid Hierarchical RL (H^2RL), introduces a logical option-based pretraining strategy to steer the learning policy away from short-term reward loops and toward goal-directed behavior while allowing the final policy to be refined via standard environment interaction. Empirically, we show that this approach consistently improves long-horizon decision-making and yields agents that outperform strong neural, symbolic, and neuro-symbolic baselines.

Analogously to classical computers, quantum processors exhibit side channels that may give attackers access to potentially proprietary algorithms. We identify and exploit a previously unexplored side channel in trapped-ion quantum processors that arises from the radio-frequency (RF) signals used to modulate lasers for ion cooling, gate execution, and readout. In these quantum processors, acousto-optical modulators (AOMs) imprint phase and frequency modulations onto laser fields interacting with the ions to implement individual and collective unitaries. The AOMs are driven by strong RF signals, a fraction of which leaks out of the device. We discuss general strategies to exploit this side channel and demonstrate how to detect RF leakage from a state-of-the-art qudit-based quantum processor using off-the-shelf components. From this data, we extract pulse characteristics of single-ion and entangling gates, thereby implementing a proof-of-principle exploitation of the novel attack vector. Finally, we outline ways to mitigate the information leakage through the presented side channel.

Understanding how neural networks transform inputs into outputs is crucial for interpreting and manipulating their behavior. Most existing approaches analyze internal representations by identifying hidden-layer activation patterns correlated with human-interpretable concepts. Here we take a direct approach to examine how hidden neurons act to drive network outputs. We introduce CODEC (Contribution Decomposition), a method that uses sparse autoencoders to decompose network behavior into sparse motifs of hidden-neuron contributions, revealing causal processes that cannot be determined by analyzing activations alone. Applying CODEC to benchmark image-classification networks, we find that contributions grow in sparsity and dimensionality across layers and, unexpectedly, that they progressively decorrelate positive and negative effects on network outputs. We further show that decomposing contributions into sparse modes enables greater control and interpretation of intermediate layers, supporting both causal manipulations of network output and human-interpretable visualizations of distinct image components that combine to drive that output. Finally, by analyzing state-of-the-art models of neural activity in the vertebrate retina, we demonstrate that CODEC uncovers combinatorial actions of model interneurons and identifies the sources of dynamic receptive fields. Overall, CODEC provides a rich and interpretable framework for understanding how nonlinear computations evolve across hierarchical layers, establishing contribution modes as an informative unit of analysis for mechanistic insights into artificial neural networks.

Virtual Reality (VR) enables users to engage with capabilities beyond human limitations, but it is not always obvious how to trigger these capabilities. Taking the lens of Affordance, we believe avatar design is the key to solving this issue, which ideally should communicate its capabilities and how to activate them. To understand the current practice, we selected eight capabilities across four categories and invited twelve professional designers to design avatars that communicate the capabilities and their corresponding interactions. From the resulting designs, we formed 16 guidelines to provide general and category-specific recommendations. Then, we validated these guidelines by letting two groups of twelve participants design avatars with and without guidelines. Participants rated the guidelines' clarity and usefulness highly. External judges confirmed that avatars designed with the guidelines were more intuitive in conveying the capabilities and interaction methods. Finally, we demonstrated the applicability of the guidelines in avatar design for four VR applications.

Hierarchical time-series forecasting is essential for demand prediction across various industries. While machine learning models have obtained significant accuracy and scalability on such forecasting tasks, the interpretability of their predictions, informed by application, is still largely unexplored. To bridge this gap, we introduce a novel interpretability method for large hierarchical probabilistic time-series forecasting, adapting generic interpretability techniques while addressing challenges associated with hierarchical structures and uncertainty. Our approach offers valuable interpretative insights in response to real-world industrial supply chain scenarios, including 1) the significance of various time-series within the hierarchy and external variables at specific time points, 2) the impact of different variables on forecast uncertainty, and 3) explanations for forecast changes in response to modifications in the training dataset. To evaluate the explainability method, we generate semi-synthetic datasets based on real-world scenarios of explaining hierarchical demands for over ten thousand products at a large chemical company. The experiments showed that our explainability method successfully explained state-of-the-art industrial forecasting methods with significantly higher explainability accuracy. Furthermore, we provide multiple real-world case studies that show the efficacy of our approach in identifying important patterns and explanations that help stakeholders better understand the forecasts. Additionally, our method facilitates the identification of key drivers behind forecasted demand, enabling more informed decision-making and strategic planning. Our approach helps build trust and confidence among users, ultimately leading to better adoption and utilization of hierarchical forecasting models in practice.

Just-in-time (JIT) compilers are key components for many popular programming languages with managed runtimes (e.g., Java and JavaScript). JIT compilers perform optimizations and generate native code at runtime based on dynamic profiling data, to improve the execution performance of the running application. Like other software systems, JIT compilers might have software bugs, and prior work has developed a number of automated techniques for detecting functional bugs (i.e., generated native code does not semantically match that of the original code). However, no prior work has targeted JIT compiler performance bugs, which can cause significant performance degradation while an application is running. These performance bugs are challenging to detect due to the complexity and dynamic nature of JIT compilers. In this paper, we present the first work on demystifying JIT performance bugs. First, we perform an empirical study across four popular JIT compilers for Java and JavaScript. Our manual analysis of 191 bug reports uncovers common triggers of performance bugs, patterns in which these bugs manifest, and their root causes. Second, informed by these insights, we propose layered differential performance testing, a lightweight technique to automatically detect JIT compiler performance bugs, and implement it in a tool called Jittery. We incorporate practical optimizations into Jittery such as test prioritization, which reduces testing time by 92.40% without compromising bug-detection capability, and automatic filtering of false-positives and duplicates, which substantially reduces manual inspection effort. Using Jittery, we discovered 12 previously unknown performance bugs in the Oracle HotSpot and Graal JIT compilers, with 11 confirmed and 6 fixed by developers.

Accurate and adaptive dynamic models are critical for underwater vehicle-manipulator systems where hydrodynamic effects induce time-varying parameters. This paper introduces a novel uncertainty-aware adaptive dynamics model framework that remains linear in lumped vehicle and manipulator parameters, and embeds convex physical consistency constraints during online estimation. Moving horizon estimation is used to stack horizon regressors, enforce realizable inertia, damping, friction, and hydrostatics, and quantify uncertainty from parameter evolution. Experiments on a BlueROV2 Heavy with a 4-DOF manipulator demonstrate rapid convergence and calibrated predictions. Manipulator fits achieve R2 = 0.88 to 0.98 with slopes near unity, while vehicle surge, heave, and roll are reproduced with good fidelity under stronger coupling and noise. Median solver time is approximately 0.023 s per update, confirming online feasibility. A comparison against a fixed parameter model shows consistent reductions in MAE and RMSE across degrees of freedom. Results indicate physically plausible parameters and confidence intervals with near 100% coverage, enabling reliable feedforward control and simulation in underwater environments.

Next-generation autonomous vehicles (AVs) rely on large volumes of multisource and multimodal ($M^2$) data to support real-time decision-making. In practice, data quality (DQ) varies across sources and modalities due to environmental conditions and sensor limitations, yet AV research has largely prioritized algorithm design over DQ analysis. This work focuses on redundancy as a fundamental but underexplored DQ issue in AV datasets. Using the nuScenes and Argoverse 2 (AV2) datasets, we model and measure redundancy in multisource camera data and multimodal image-LiDAR data, and evaluate how removing redundant labels affects the YOLOv8 object detection task. Experimental results show that selectively removing redundant multisource image object labels from cameras with shared fields of view improves detection. In nuScenes, mAP${50}$ gains from $0.66$ to $0.70$, $0.64$ to $0.67$, and from $0.53$ to $0.55$, on three representative overlap regions, while detection on other overlapping camera pairs remains at the baseline even under stronger pruning. In AV2, $4.1$-$8.6\%$ of labels are removed, and mAP${50}$ stays near the $0.64$ baseline. Multimodal analysis also reveals substantial redundancy between image and LiDAR data. These findings demonstrate that redundancy is a measurable and actionable DQ factor with direct implications for AV performance. This work highlights the role of redundancy as a data quality factor in AV perception and motivates a data-centric perspective for evaluating and improving AV datasets. Code, data, and implementation details are publicly available at: https://github.com/yhZHOU515/RedundancyAD

We introduce SurgFormer, a multiresolution gated transformer for data driven soft tissue simulation on volumetric meshes. High fidelity biomechanical solvers are often too costly for interactive use, so we train SurgFormer on solver generated data to predict nodewise displacement fields at near real time rates. SurgFormer builds a fixed mesh hierarchy and applies repeated multibranch blocks that combine local message passing, coarse global self attention, and pointwise feedforward updates, fused by learned per node, per channel gates to adaptively integrate local and long range information while remaining scalable on large meshes. For cut conditioned simulation, resection information is encoded as a learned cut embedding and provided as an additional input, enabling a unified model for both standard deformation prediction and topology altering cases. We also introduce two surgical simulation datasets generated under a unified protocol with XFEM based supervision: a cholecystectomy resection dataset and an appendectomy manipulation and resection dataset with cut and uncut cases. To our knowledge, this is the first learned volumetric surrogate setting to study XFEM supervised cut conditioned deformation within the same volumetric pipeline as standard deformation prediction. Across diverse baselines, SurgFormer achieves strong accuracy with favorable efficiency, making it a practical backbone for both tasks. {Code, data, and project page: \href{https://mint-vu.github.io/SurgFormer/}{available here}}

Conversational generative AI is rapidly entering healthcare, where general-purpose models must integrate heterogeneous patient signals and support diverse interaction styles while producing clinically meaningful outputs. In respiratory care, non-invasive audio, such as recordings captured via mobile microphones, enables scalable screening and longitudinal monitoring, but the heterogeneity challenge is particularly acute: recordings vary widely across devices, environments, and acquisition protocols, and questions span multiple intents and question formats. Existing biomedical audio-language QA systems are typically monolithic, without any specialization mechanisms for tackling diverse respiratory corpora and query intents. They are also only validated in limited settings, leaving it unclear how reliably they handle the shifts encountered in real-world settings. To address these limitations, we introduce RAMoEA-QA, a hierarchically routed generative model for respiratory audio question answering that unifies multiple question types and supports both discrete and continuous targets within a single multimodal system. RAMoEA-QA applies two-stage conditional specialization: an Audio Mixture-of-Experts routes each recording to a suitable pre-trained audio encoder, and a Language Mixture-of-Adapters selects a LoRA adapter on a shared frozen LLM to match the query intent and answer format. By specializing both acoustic representations and generation behaviour per example, RAMoEA-QA consistently outperforms strong baselines and routing ablations with minimal parameter overhead, improving in-domain test accuracy to 0.72 (vs. 0.61 and 0.67 for state-of-the-art baselines) and exhibiting the strongest generalization for diagnosis under domain, modality, and task shifts.

In our previous work [2], we introduced a hardware- and power-efficient architecture for hybrid digital-analog (HDA) multiuser MIMO (MU-MIMO) based on stacking identical basic modules. Each module consists of a small active multi-antenna feeder (AMAF) placed in the near field of a larger reflective intelligent surface (RIS). Each AMAF is driven by one RF chain and conveys one spatial stream, achieving a multiplexing gain of $K$ with $K$ stacked modules. While [2] focused on module design and efficiency compared to active arrays, performance was evaluated only under pure line-of-sight (LOS) conditions. This work extends our approach in several ways. First, we propose a simple, pragmatic method for designing phase-only flat-top beams for the AMAF-RIS module, enabling wide angular coverage with low ripple and sidelobes. This design supports hierarchical beamforming codebooks for efficient beam acquisition. Second, we evaluate MU-MIMO performance under realistic mmWave multipath channels including both LOS and non-LOS (NLOS) components modeled using a 3D von Mises-Fisher distribution. We propose a low-complexity HDA MU-MIMO framework with: user-beam association via standard beam acquisition; dynamic user grouping (one user per beam); effective baseband MIMO channel estimation using 3GPP-compliant pilots; and downlink transmission with zero-forcing precoding under per-antenna power constraints. Results show high spectral efficiency and multiplexing gain while preserving hardware simplicity and power efficiency. Crucially, the approach is fully compliant with 3GPP 5GNR beam acquisition and sounding reference signaling mechanisms.

Device logs are essential for forensic investigations, enterprise monitoring, and fraud detection; however, they often leak personally identifiable information (PII) when exported for third-party analysis. Existing approaches either fail to minimize PII exposure across all stages of log collection and analysis or sacrifice data fidelity, resulting in less effective analysis. We present Proteus, a privacy-preserving device logging framework that enables forensic analysis without disclosing plaintext PII or compromising fidelity, even when facing adversaries with access to multiple snapshots of the log files. To achieve this, Proteus proposes a two-layer scheme that employs keyed-hash pseudonymization of PII fields and time-rotating encryption with ratcheted ephemeral keys to prevent multi-snapshot correlation. For controlled sharing, clients export ratchet states that grant time-bounded access, permitting decryption of pseudonymized tokens that enable linkage and timeline reconstruction without exposing the underlying PII. Subsequent ratchet rotations ensure forward secrecy, while DICE-based attestation authenticates device provenance. We implement Proteus as a transparent extension to Android's logcat and evaluate it across three generations of hardware. Our results demonstrate a median latency of 0.2 ms per message and an average per-PII-field size overhead of only 97.1 bytes.

Temporal task structure is fundamental for bimanual manipulation: a robot must not only know that one action precedes or overlaps another, but also when each action should occur and how long it should take. While symbolic temporal relations enable high-level reasoning about task structure and alternative execution sequences, concrete timing parameters are equally essential for coordinating two hands at the execution level. Existing approaches address these two levels in isolation, leaving a gap between high-level task planning and low-level movement synchronization. This work presents an approach for learning both symbolic and subsymbolic temporal task constraints from human demonstrations and deriving executable, temporally parametrized plans for bimanual manipulation. Our contributions are (i) a 3-dimensional representation of timings between two actions with methods based on multivariate Gaussian Mixture Models to represent temporal relationships between actions on a subsymbolic level, (ii) a method based on the Davis-Putnam-Logemann-Loveland (DPLL) algorithm that finds and ranks all contradiction-free assignments of Allen relations to action pairs, representing different modes of a task, and (iii) an optimization-based planning system that combines the identified symbolic and subsymbolic temporal task constraints to derive temporally parametrized plans for robot execution. We evaluate our approach on several datasets, demonstrating that our method generates temporally parametrized plans closer to human demonstrations than the most characteristic demonstration baseline.

In this paper, we propose an adaptive data-driven min-max model predictive control (MPC) scheme for discrete-time linear time-varying (LTV) systems. We assume that prior knowledge of the system dynamics and bounds on the variations are known, and that the states are measured online. Starting from an initial state-feedback gain derived from prior knowledge, the algorithm updates the state-feedback gain using online input-state data. To this end, a semidefinite program (SDP) is solved to minimize an upper bound on the infinite-horizon optimal cost and to derive a corresponding state-feedback gain. We prove that the resulting closed-loop system is exponentially stabilized and satisfies the constraints. Further, we extend the proposed scheme to LTV systems with process noise. The resulting closed-loop system is shown to be robustly stabilized to a robust positive invariant (RPI) set. Finally, the proposed methods are demonstrated by numerical simulations.

Negation is a fundamental linguistic operator, yet it remains inadequately modeled in diffusion-based generative systems. In this work, we present a formal treatment of linguistic negation in diffusion-based generative models by modeling it as a structured feasibility constraint on semantic guidance within diffusion dynamics. Rather than introducing heuristics or retraining model parameters, we reinterpret classifier-free guidance as defining a semantic update direction and enforce negation by projecting the update onto a convex constraint set derived from linguistic structure. This novel formulation provides a unified framework for handling diverse negation phenomena, including object absence, graded non-inversion semantics, multi-negation composition, and scope-sensitive disambiguation. Our approach is training-free, compatible with pretrained diffusion backbones, and naturally extends from image generation to temporally evolving video trajectories. In addition, we introduce a structured negation-centric benchmark suite that isolates distinct linguistic failure modes in generative systems, to further research in this area. Experiments demonstrate that our method achieves robust negation compliance while preserving visual fidelity and structural coherence, establishing the first unified formulation of linguistic negation in diffusion-based generative models beyond representation-level evaluation.

Diffuse direct time-of-flight LiDARs report per-pixel depth histograms formed by aggregating photon returns over a wide instantaneous field of view, violating the single-ray assumption behind standard LiDAR-RGB calibration. We present a simple spatial calibration procedure that estimates, for each diffuse LiDAR pixel, its footprint (effective support region) and relative spatial sensitivity in a co-located RGB image plane. Using a scanned retroreflective patch with background subtraction, we recover per-pixel response maps that provide an explicit LiDAR-to-RGB correspondence for cross-modal alignment and fusion. We demonstrate the method on the ams OSRAM TMF8828.

When humans perceive the world, they naturally integrate multiple audio-visual tasks within dynamic, real-world scenes. However, current works such as event localization, parsing, segmentation and question answering are mostly explored individually, making it challenging to comprehensively understand complex audio-visual scenes and explore inter-task relationships. Hence, we propose \textbf{AV-Unified}, a unified framework that enables joint learning across a wide range of audio-visual scene understanding tasks. AV-Unified standardizes the diverse input-output formats of each task and incorporates a multi-scale spatiotemporal perception network to effectively capture audio-visual associations. Specifically, we unify the inputs and outputs of all supported tasks by converting them into sequences of discrete tokens, establishing a shared representation that allows a single architecture to be trained jointly across heterogeneous varied datasets. Considering the varying temporal granularity of audio-visual events, a multi-scale temporal perception module is designed to capture key cues. Meanwhile, to overcome the lack of auditory supervision in the visual domain, we design a cross-modal guidance-based spatial perception module that models spatial audio-visual associations. Furthermore, task-specific text prompts are employed to enhance the model's adaptability and task-awareness. Extensive experiments on benchmark datasets (e.g., AVE, LLP, MUSIC-AVQA, VGG-SS and AVS) demonstrate the effectiveness of AV-Unified across temporal, spatial, and spatiotemporal tasks.

We present a rolling and jumping underactuated monopedal robot designed to explore multimodal locomotion on low-gravity bodies. It uses only two reaction wheels to control its spatial orientation with two controllers: a balancing controller which can aim the robot's jump direction on the ground, and an aerial reorientation controller which can aim the robot's leg for landing after flight. We demonstrate rolling, targeted jumping and landing, and self-righting using only three actuators total, keeping system size to 0.33m and 1.25kg. Simple switching between locomotion modes enables the system to deal with differing landscapes and environmental conditions.

Explainable AI (XAI) has become essential in computer vision to make the decision-making processes of deep learning models transparent. However, current visual explanation (XAI) methods face a critical trade-off between the high fidelity of architecture-specific methods and the broad applicability of universal ones. This often results in abstract or fragmented explanations and makes it difficult to compare explanatory power across diverse model families, such as CNNs and Transformers. This paper introduces the Self-Confidence and Analysis Networks (SCAN), a novel universal framework that overcomes these limitations for both convolutional neural network and transformer architectures. SCAN utilizes an AutoEncoder-based approach to reconstruct features from a model's intermediate layers. Guided by the Information Bottleneck principle, it generates a high-resolution Self-Confidence Map that identifies information-rich regions. Extensive experiments on diverse architectures and datasets demonstrate that SCAN consistently achieves outstanding performance on various quantitative metrics such as AUC-D, Negative AUC, Drop%, and Win%. Qualitatively, it produces significantly clearer, object-focused explanations than existing methods. By providing a unified framework that is both architecturally universal and highly faithful, SCAN enhances model transparency and offers a more reliable tool for understanding the decision-making processes of complex neural networks.

Orofacial clefts are among the most common congenital craniofacial abnormalities, yet accurate prenatal detection remains challenging due to the scarcity of experienced specialists and the relative rarity of the condition. Early and reliable diagnosis is essential to enable timely clinical intervention and reduce associated morbidity. Here we show that an artificial intelligence system, trained on over 45,139 ultrasound images from 9,215 fetuses across 22 hospitals, can diagnose fetal orofacial clefts with sensitivity and specificity exceeding 93% and 95% respectively, matching the performance of senior radiologists and substantially outperforming junior radiologists. When used as a medical copilot, the system raises junior radiologists' sensitivity by more than 6%. Beyond direct diagnostic assistance, the system also accelerates the development of clinical expertise. A pilot study involving 24 radiologists and trainees demonstrated that the model can improve the expertise development for rare conditions. This dual-purpose approach offers a scalable solution for improving both diagnostic accuracy and specialist training in settings where experienced radiologists are scarce.

Robotic harvesting in dense crop canopies requires effective interventions that depend not only on geometry, but also on explicit, direction-conditioned relations identifying which organs obstruct a target fruit. We present SG-DOR (Scene Graphs with Direction-Conditioned Occlusion Reasoning), a relational framework that, given instance-segmented organ point clouds, infers a scene graph encoding physical attachments and direction-conditioned occlusion. We introduce an occlusion ranking task for retrieving and ranking candidate leaves for a target fruit and approach direction, and propose a direction-aware graph neural architecture with per-fruit leaf-set attention and union-level aggregation. Experiments on a multi-plant synthetic pepper dataset show improved occlusion prediction (F1=0.73, NDCG@3=0.85) and attachment inference (edge F1=0.83) over strong ablations, yielding a structured relational signal for downstream intervention planning.

While diffusion models have revolutionized visual content generation, their rapid adoption has underscored the critical need to investigate vulnerabilities, e.g., to backdoor attacks. In multimodal diffusion models, it is natural to expect that attacking multiple modalities simultaneously (e.g., text and image) would yield complementary effects and strengthen the overall backdoor. In this paper, we challenge this assumption by investigating the phenomenon of Backdoor Modality Collapse, a scenario where the backdoor mechanism degenerates to rely predominantly on a subset of modalities, rendering others redundant. To rigorously quantify this behavior, we introduce two novel metrics: Trigger Modality Attribution (TMA) and Cross-Trigger Interaction (CTI). Through extensive experiments across diverse training configurations in multimodal conditional diffusion, we consistently observe a ``winner-takes-all'' dynamic in backdoor behavior. Our results reveal that (1) attacks often collapse into subset-modality dominance, and (2) cross-modal interaction is negligible or even negative, contradicting the intuition of synergistic vulnerability. These findings highlight a critical blind spot in current assessments, suggesting that high attack success rates often mask a fundamental reliance on a subset of modalities. This establishes a principled foundation for mechanistic analysis and future defense development.

Strong semantic representations improve the convergence and generation quality of diffusion and flow models. Existing approaches largely rely on external models, which require separate training, operate on misaligned objectives, and exhibit unexpected scaling behavior. We argue that this dependence arises from the model's training objective, which poses a denoising task with little incentive to learn semantic representations. We introduce Self-Flow: a self-supervised flow matching paradigm that integrates representation learning within the generative framework. Our key mechanism, Dual-Timestep Scheduling, applies heterogeneous noise levels across tokens, creating an information asymmetry that forces the model to infer missing information from corrupted inputs. This drives learning strong representations alongside generative capabilities without external supervision. Our method generalizes across modalities and enables multi-modal training while following expected scaling laws, achieving superior image, video, and audio generation.

Automatic speech recognition (ASR) has benefited from advances in pretrained speech and language models, yet most systems remain constrained to monolingual settings and short, isolated utterances. While recent efforts in context-aware ASR show promise, two key challenges persist: limited multilingual support and the absence of principled alignment between speech and contextual representations. In this paper, we introduce a context-aware multilingual ASR framework that supports diverse languages and accents while preserving the modularity of pretrained models. Our approach combines a frozen speech encoder and a decoder-only language model via a lightweight projection module, allowing structured context prompts, including dialogue history and biasing words, to guide transcription. To improve interaction between speech and context, we employ a contrastive learning objective that aligns their representations in a shared embedding space. Evaluations on over 1,500 hours of real-world conversational speech across 11 languages and 5 English dialects show that contextual input consistently improves recognition quality. Contrastive alignment provides additional gains when applied to different context types, with an overall performance gain of over 5%. These results highlight the importance of both contextual modeling and cross-modal alignment in multilingual ASR.

Recent advances in multimodal Retrieval-Augmented Generation (RAG) enable Large Language Models (LLMs) to analyze enterprise spreadsheet workbooks containing millions of cells, cross-sheet dependencies, and embedded visual artifacts. However, state-of-the-art approaches exclude critical context through single-pass retrieval, lose data resolution through compression, and exceed LLM context windows through naive full-context injection, preventing reliable multi-step reasoning over complex enterprise workbooks. We introduce Beyond Rows to Reasoning (BRTR), a multimodal agentic framework for spreadsheet understanding that replaces single-pass retrieval with an iterative tool-calling loop, supporting end-to-end Excel workflows from complex analysis to structured editing. Supported by over 200 hours of expert human evaluation, BRTR achieves state-of-the-art performance across three frontier spreadsheet understanding benchmarks, surpassing prior methods by 25 percentage points on FRTR-Bench, 7 points on SpreadsheetLLM, and 32 points on FINCH. We evaluate five multimodal embedding models, identifying NVIDIA NeMo Retriever 1B as the top performer for mixed tabular and visual data, and vary nine LLMs. Ablation experiments confirm that the planner, retrieval, and iterative reasoning each contribute substantially, and cost analysis shows GPT-5.2 achieves the best efficiency-accuracy trade-off. Throughout all evaluations, BRTR maintains full auditability through explicit tool-call traces.