Standard transformers entangle all computation in a single residual stream, obscuring which components perform which functions. We introduce the Dual-Stream Transformer, which decomposes the residual stream into two functionally distinct components: a token stream updated by attention and a context stream updated by feed-forward networks. Information flow between attention heads is controlled through a hierarchy of mixing strategies, from fully independent (maximum interpretability) to dense (standard transformer behavior). This design exposes a tunable tradeoff between interpretability and performance. We measure this tradeoff on language modeling tasks at 29M parameters. Fully independent head mixing increases validation loss by 8\% relative to dense baselines. The recommended Kronecker mixing strategy, which permits scalar communication between heads while preserving within-head structure, costs only 2.5\%. All configurations maintain functional generation under attention amplification (scaling logits by factors up to 16 at inference time), with degradation ranging from 16\% to 27\%. This robustness suggests the architectures learn discrete algorithms that operate independently of soft probabilistic mixing. The architecture provides a foundation for interpretable language models where internal structure is exposed by design. \footnote{This work was partially supported by DARPA Contract HR001125C0302.}

We present SLNet, a lightweight backbone for 3D point cloud recognition designed to achieve strong performance without the computational cost of many recent attention, graph, and deep MLP based models. The model is built on two simple ideas: NAPE (Nonparametric Adaptive Point Embedding), which captures spatial structure using a combination of Gaussian RBF and cosine bases with input adaptive bandwidth and blending, and GMU (Geometric Modulation Unit), a per channel affine modulator that adds only 2D learnable parameters. These components are used within a four stage hierarchical encoder with FPS+kNN grouping, nonparametric normalization, and shared residual MLPs. In experiments, SLNet shows that a very small model can still remain highly competitive across several 3D recognition tasks. On ModelNet40, SLNet-S with 0.14M parameters and 0.31 GFLOPs achieves 93.64% overall accuracy, outperforming PointMLP-elite with 5x fewer parameters, while SLNet-M with 0.55M parameters and 1.22 GFLOPs reaches 93.92%, exceeding PointMLP with 24x fewer parameters. On ScanObjectNN, SLNet-M achieves 84.25% overall accuracy within 1.2 percentage points of PointMLP while using 28x fewer parameters. For large scale scene segmentation, SLNet-T extends the backbone with local Point Transformer attention and reaches 58.2% mIoU on S3DIS Area 5 with only 2.5M parameters, more than 17x fewer than Point Transformer V3. We also introduce NetScore+, which extends NetScore by incorporating latency and peak memory so that efficiency can be evaluated in a more deployment oriented way. Across multiple benchmarks and hardware settings, SLNet delivers a strong overall balance between accuracy and efficiency. Code is available at: https://github.com/m-saeid/SLNet.

Gradient boosting still dominates Transformers on tabular benchmarks. Our tokenizer uses a deliberately simplistic discretized vocabulary so we can highlight how even basic tokenization unlocks the power of attention on tabular features, yet it already outperforms tuned gradient boosting when combined with Gaussian smoothing. Our solution discretizes environmental context while smoothing labels with adaptive Gaussians, yielding calibrated PDFs. On 600K entities (5M training examples) we outperform tuned XGBoost by 10.8% (35.94s vs 40.31s median MAE) and achieve KS=0.0045 with the adaptive-sigma checkpoint selected to minimize KS rather than median MAE. Ablations confirm architecture matters: losing sequential ordering costs about 2.0%, dropping the time-delta tokens costs about 1.8%, and a stratified calibration analysis reveals where miscalibration persists.

Large language models (LLMs) have enabled agent-based systems that aim to automate scientific research workflows. Most existing approaches focus on fully autonomous discovery, where AI systems generate research ideas, conduct analyses, and produce manuscripts with minimal human involvement. However, empirical research in economics and the social sciences poses additional constraints: research questions must be grounded in available datasets, identification strategies require careful design, and human judgment remains essential for evaluating economic significance. We introduce HLER (Human-in-the-Loop Economic Research), a multi-agent architecture that supports empirical research automation while preserving critical human oversight. The system orchestrates specialized agents for data auditing, data profiling, hypothesis generation, econometric analysis, manuscript drafting, and automated review. A key design principle is dataset-aware hypothesis generation, where candidate research questions are constrained by dataset structure, variable availability, and distributional diagnostics, reducing infeasible or hallucinated hypotheses. HLER further implements a two-loop architecture: a question quality loop that screens and selects feasible hypotheses, and a research revision loop where automated review triggers re-analysis and manuscript revision. Human decision gates are embedded at key stages, allowing researchers to guide the automated pipeline. Experiments on three empirical datasets show that dataset-aware hypothesis generation produces feasible research questions in 87% of cases (versus 41% under unconstrained generation), while complete empirical manuscripts can be produced at an average API cost of $0.8-$1.5 per run. These results suggest that Human-AI collaborative pipelines may provide a practical path toward scalable empirical research.

Medical Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities across diverse healthcare tasks. However, current post-training strategies, such as supervised fine-tuning and reinforcement learning, heavily depend on substantial annotated data while overlooking the potential of unlabeled test data for model enhancement. This limitation becomes particularly pronounced in medical domains, where acquiring extensive labeled medical data is difficult due to the strict data sensitivity and annotation complexity. Moreover, leveraging test data poses challenges in generating reliable supervision signals from unlabeled samples and maintaining stable self-evolution. To address these limitations, we propose Med-Evo, the first self-evolution framework for medical MLLMs that utilizes label-free reinforcement learning to promote model performance without requiring additional labeled data. Our framework introduces two key innovations: $1)$ Feature-driven Pseudo Labeling (FPL) that identifies semantic centroids from all heterogeneous candidate responses to select pseudo labels in each rollout, and $2)$ Hard-Soft Reward (HSR) that combines exact match with token-level assessment and semantic similarity to provide hierarchical reward. Experiments on three medical VQA benchmarks and two base MLLMs show clear advantages of our approach over SOTA methods, with significant improvements of 10.43\% accuracy and 4.68\% recall on the SLAKE dataset using Qwen2.5-VL, showing the effectiveness of our method.

Modern robotic systems rely on hierarchical control, where a high-level "Brain" (Python) directs a lower-level "Spine" (C++ real-time controller). Despite its necessity, this hierarchy makes it difficult for the Brain to completely rewrite the Spine's immutable control logic, consequently inhibiting fundamental adaptation for different tasks and environments. Conventional approaches require complex middleware, proprietary hardware, or sacrifice real-time performance. We present LITHE (Linux Isolated Threading for Hierarchical Execution), a lightweight software architecture that collapses the robot control hierarchy onto a commodity single-board computer (Raspberry Pi 4B with pi3hat), while maintaining safe frequency decoupling between the Brain and Spine. LITHE integrates strict CPU isolation (isolcpus), lock-free inter-process communication (IPC), and pipelined execution to meet high-frequency deadlines with minimal jitter. By adding multi-threaded dynamic linking, LITHE enables a Python-based Brain to dynamically evolve the logic of a 1kHz C++ Spine without interruption. We validate "functional real-time" system performance with worst-case execution time (WCET) < 100 $μ$s and maximum release jitter (MRJ) < 4 $μ$s under heavy load. We demonstrate a novel application where a large language model (LLM) supervisor performs online system identification to evolve a real-time controller on-the-fly, without interrupting the 1 kHz control loop. In essence, LITHE eliminates the "immutable compiled code" bottleneck for best-effort Brains to synthesize and inject completely new control laws into the real-time Spine. This bridges a critical gap between high-level AI and low-level real-time control to unlock continuous real-time evolution of embodied intelligence in safe, human-in-the-loop systems.

Monocular 3D animal reconstruction is challenging due to complex articulation, self-occlusion, and fine-scale details such as fur. Existing methods often produce distorted geometry and inconsistent textures due to the lack of articulated 3D supervision and limited availability of back-view images in 2D datasets, which makes reconstructing unobserved regions particularly difficult. To address these limitations, we propose DogWeave, a model-based framework for reconstructing high-fidelity 3D canine models from a single RGB image. DogWeave improves geometry by refining a coarsely-initiated parametric mesh into a detailed SDF representation through multi-view normal field optimization using diffusion-enhanced normals. It then generates view-consistent textures through conditional partial inpainting guided by structure and style cues, enabling realistic reconstruction of unobserved regions. Using only about 7,000 dog images processed via our 2D pipeline for training, DogWeave produces complete, realistic 3D models and outperforms state-of-the-art single image to 3d reconstruction methods in both shape accuracy and texture realism for canines.

We study the problem of state representation learning for control from partial and potentially high-dimensional observations. We approach this problem via cost-driven state representation learning, in which we learn a dynamical model in a latent state space by predicting cumulative costs. In particular, we establish finite-sample guarantees on finding a near-optimal representation function and a near-optimal controller using the learned latent model for infinite-horizon time-invariant Linear Quadratic Gaussian (LQG) control. We study two approaches to cost-driven representation learning, which differ in whether the transition function of the latent state is learned explicitly or implicitly. The first approach has also been investigated in Part I of this work, for finite-horizon time-varying LQG control. The second approach closely resembles MuZero, a recent breakthrough in empirical reinforcement learning, in that it learns latent dynamics implicitly by predicting cumulative costs. A key technical contribution of this Part II is to prove persistency of excitation for a new stochastic process that arises from the analysis of quadratic regression in our approach, and may be of independent interest.

Graphical user interface (GUI)-based mobile agents automate digital tasks on mobile devices by interpreting natural-language instructions and interacting with the screen. While recent methods apply reinforcement learning (RL) to train vision-language-model(VLM) agents in interactive environments with a primary focus on performance, generalization remains underexplored due to the lack of standardized benchmarks and open-source RL systems. In this work, we formalize the problem as a Contextual Markov Decision Process (CMDP) and introduce \textbf{AndroidWorld-Generalization}, a benchmark with three increasingly challenging regimes for evaluating zero-shot generalization to unseen task instances, templates, and applications. We further propose an RL training system that integrates Group Relative Policy Optimization (GRPO) with a scalable rollout collection system, consisting of containerized infrastructure and asynchronous execution % , and error recovery to support reliable and efficient training. Experiments on AndroidWorld-Generalization show that RL enables a 7B-parameter VLM agent to surpass supervised fine-tuning baselines, yielding a 26.1\% improvement on unseen instances but only limited gains on unseen templates (15.7\%) and apps (8.3\%), underscoring the challenges of generalization. As a preliminary step, we demonstrate that few-shot adaptation at test-time improves performance on unseen apps, motivating future research in this direction. To support reproducibility and fair comparison, we open-source the full RL training system, including the environment, task suite, models, prompt configurations, and the underlying infrastructure \footnote{https://github.com/zihuanjiang/AndroidWorld-Generalization}.

Image Super-Resolution (SR) aims to reconstruct high-resolution images from degraded low-resolution inputs. While diffusion-based SR methods offer powerful generative capabilities, their performance heavily depends on how semantic priors are structured and integrated into the generation process. Existing approaches often rely on entangled or coarse-grained priors that mix global layout with local details, or conflate structural and textural cues, thereby limiting semantic controllability and interpretability. In this work, we propose DTPSR, a novel diffusion-based SR framework that introduces disentangled textual priors along two complementary dimensions: spatial hierarchy (global vs. local) and frequency semantics (low- vs. high-frequency). By explicitly separating these priors, DTPSR enables the model to simultaneously capture scene-level structure and object-specific details with frequency-aware semantic guidance. The corresponding embeddings are injected via specialized cross-attention modules, forming a progressive generation pipeline that reflects the semantic granularity of visual content, from global layout to fine-grained textures. To support this paradigm, we construct DisText-SR, a large-scale dataset containing approximately 95,000 image-text pairs with carefully disentangled global, low-frequency, and high-frequency descriptions. To further enhance controllability and consistency, we adopt a multi-branch classifier-free guidance strategy with frequency-aware negative prompts to suppress hallucinations and semantic drift. Extensive experiments on synthetic and real-world benchmarks show that DTPSR achieves high perceptual quality, competitive fidelity, and strong generalization across diverse degradation scenarios.

Micro-scale continuum robots face significant limitations in achieving three-dimensional contact force perception, primarily due to structural miniaturization, nonlinear mechanical, and sensor integration. To overcome these limitations, this paper introduces a novel proprioception method for cable-driven continuum robots based on proximal-integrated force sensing (i.e., cable tension and six-axis force/torque (F/T) sensor), inspired by the tendon-joint collaborative sensing mechanism of the finger. By integrating biomechanically inspired design principles with nonlinear modeling, the proposed method addresses the challenge of force perception (including the three-dimensional contact force and the location of the contact point) and shape estimation in micro-scale continuum robots. First, a quasi-bionic mapping between human tissues/organs and robot components is established, enabling the transfer of the integrated sensing strategy of tendons, joints, and neural feedback to the robotic system. Second, a multimodal perception strategy is developed based on the structural constraints inherent to continuum robots. The complex relationships among mechanical and material nonlinearities, robot motion states, and contact forces are formulated as an optimization problem to reduce the perception complexity. Finally, experimental validation demonstrates the effectiveness of the proposed method. This work lays the foundation for developing safer and smarter continuum robots, enabling broader clinical adoption in complex environments.

Centipede-like robots offer unique locomotion advantages due to their small cross-sectional area for accessing confined spaces, and their redundant legs enhance robustness in cluttered environments such as search-and-rescue and pipe inspection. However, elongated robots are particularly vulnerable to tipping over when climbing large obstacles, making reliable self-righting essential for field deployment. Self-righting strategies for elongate, multi-legged systems remain poorly understood. In this study, we conduct a comparative biomechanics and robophysical investigation to address three key questions: (1) What self-righting strategies are effective for elongate, many-legged systems? (2) How should these strategies depend on morphological parameters such as leg length and leg number? (3) Is there a morphological limit beyond which reliable self-righting becomes infeasible? We compare two biological exemplars: Scolopendra subspinipes (short legs) and Scutigera coleoptrata (house centipedes with long legs). Scolopendra subspinipes reliably self-rights both during aerial phases and through ground-assisted self-righting, whereas house centipedes rely predominantly on aerial reorientation and struggle to generate effective self-righting torques during ground contact. Motivated by these observations, we construct a parameterized space of bio-inspired self-righting strategies and develop an elongate robot with adjustable leg lengths. Systematic experiments reveal that increasing leg length necessitates a shift in control strategy to prevent torque over-concentration in mid-body actuators, and we identify a critical limb-length threshold above which robust self-righting becomes challenging. These results establish morphology-strategy coupling principles for self-righting in elongate robots and provide design guidelines for centipede-like systems operating in uncertain terrain.

Large language model-based deep research agents have been increasingly popular for addressing long-horizon information-seeking tasks, but they often incur high end-to-end latency due to extensive reasoning and frequent tool use. Speculation frameworks aim to reduce latency by overlapping action execution with reasoning; however, existing approaches typically rely on uniform speculation strategies and strict action matching, which limits inference speedups and robustness. In this work, we revisit the speculate-verify paradigm for deep research agents through the lens of action heterogeneity. We show that \textit{Search} and \textit{Visit} actions exhibit fundamentally different reasoning and model capacity requirements: entropy-based analysis reveals that Search decisions have higher uncertainty and benefit significantly from explicit reasoning, whereas Visit decisions have lower entropy and depend primarily on model capacity. Motivated by this dual-process characteristic, we propose DualSpec, a heterogeneous speculation framework equipped with a lightweight, confidence-based semantic verifier. Experiments across multiple models and benchmarks demonstrate that DualSpec achieves up to 3.28$\times$ end-to-end speedup while maintaining accuracy comparable to fully reasoning agents.

Catastrophic forgetting remains a central challenge in continual learning (CL), yet lacks a unified information-theoretic explanation for why some architectures forget catastrophically while others do not. We introduce \emph{Context Channel Capacity} ($C_\mathrm{ctx}$), the mutual information between a CL architecture's context signal and its generated parameters, and prove that zero forgetting requires $C_\mathrm{ctx} \geq H(T)$, where $H(T)$ is the task identity entropy. We establish an \emph{Impossibility Triangle} -- zero forgetting, online learning, and finite parameters cannot be simultaneously satisfied by sequential state-based learners -- and show that conditional regeneration architectures (HyperNetworks) bypass this triangle by redefining parameters as function values rather than states. We validate this framework across 8 CL methods on Split-MNIST (1,130+ experiments over 86 days, 4 seeds each), showing that $C_\mathrm{ctx}$ perfectly predicts forgetting behavior: methods with $C_\mathrm{ctx} = 0$ (NaiveSGD, EWC, SI, LwF, CFlow) exhibit catastrophic forgetting (6--97\%), while methods with $C_\mathrm{ctx} \approx 1$ (HyperNetwork) achieve zero forgetting (98.8\% ACC). We further propose \emph{Wrong-Context Probing} (P5), a practical diagnostic protocol for measuring $C_\mathrm{ctx}$, and extend the framework to CIFAR-10 via a novel \emph{Gradient Context Encoder} that closes the oracle gap from 23.3pp to 0.7pp. A systematic taxonomy of 15+ closed research directions -- including the Hebbian null result (frozen random features outperform learned features), CFlow's $θ_0$-memorizer phenomenon, and the $S_N$ symmetry barrier to column specialization -- provides the community with precisely diagnosed negative results. Our central design principle: \emph{architecture over algorithm} -- the context pathway must be structurally unbypassable.

Visual place recognition (VPR) aiming at predicting the location of an image based solely on its visual features is a fundamental task in robotics and autonomous systems. Domain variation remains one of the main challenges in VPR and is relatively unexplored. Existing VPR models attempt to achieve domain agnosticism either by training on large-scale datasets that inherently contain some domain variations, or by being specifically adapted to particular target domains. In practice, the former lacks explicit domain supervision, while the latter generalizes poorly to unseen domain shifts. This paper proposes a novel query-based domain-agnostic VPR model called QdaVPR. First, a dual-level adversarial learning framework is designed to encourage domain invariance for both the query features forming the global descriptor and the image features from which these query features are derived. Then, a triplet supervision based on query combinations is designed to enhance the discriminative power of the global descriptors. To support the learning process, we augment a large-scale VPR dataset using style transfer methods, generating various synthetic domains with corresponding domain labels as auxiliary supervision. Extensive experiments show that QdaVPR achieves state-of-the-art performance on multiple VPR benchmarks with significant domain variations. Specifically, it attains the best Recall@1 and Recall@10 on nearly all test scenarios: 93.5%/98.6% on Nordland (seasonal changes), 97.5%/99.0% on Tokyo24/7 (day-night transitions), and the highest Recall@1 across almost all weather conditions on the SVOX dataset. Our code will be released at https://github.com/shuimushan/QdaVPR.

Universal image restoration aims to recover clean images from arbitrary real-world degradations using a single inference model. Despite significant progress, existing all-in-one restoration networks do not scale to multiple degradations. As the number of degradations increases, training becomes unstable, models grow excessively large, and performance drops across both seen and unseen domains. In this work, we show that scaling universal restoration is fundamentally limited by interference across degradations during joint learning, leading to catastrophic task forgetting. To address this challenge, we introduce a unified inference pipeline with a multi-branch mixture-of-experts architecture that decomposes restoration knowledge across specialized task-adaptable experts. Our approach enables scalable learning (over sixteen degradations), adapts and generalizes robustly to unseen domains, and supports user-controllable restoration across degradations. Beyond achieving superior performance across benchmarks, this work establishes a new design paradigm for scalable and controllable universal image restoration.

Vision language action models (VLAs) are increasingly used for Physical AI, but deploying a pre-trained VLA model to unseen environments, embodiments, or tasks still requires adaptation. Parameter-efficient fine-tuning (PEFT), especially LoRA, is common for VLA policies, yet the exposed capacity knob, the rank, does not transfer uniformly: robotics transfer exhibits a higher and task-varying intrinsic rank than language fine-tuning. Small ranks suffice for LLMs (e.g., $r \in \{4, 8\}$), while spectral analyses indicate VLAs may require much larger ranks (e.g., $r \approx 128$) or near-full rank, a mismatch that worsens in multi-task settings. We present LoRA-SP (Select-Prune), a rank-adaptive fine-tuning method that replaces fixed-rank updates with input- and layer-wise capacity. LoRA-SP uses an SVD-style parameterization with a small router whose nonnegative scores act as singular values over a shared vector bank. The active set is chosen by an energy target on the cumulative squared scores $E(k) \ge η$, providing a direct link to approximation error via our spectral analysis. During training, $η$ concentrates energy on a few directions and teaches the router to rely on fewer vectors while preserving accuracy. This yields compact adapters that reduce cross-task interference and improve generalization. On four real-robot manipulation tasks collected on an unseen AgileX PiPER arm, across two VLA backbones ($π_0$ and SmolVLA), LoRA-SP matches or exceeds full fine-tuning with far fewer trainable parameters, and improves multi-task success by up to 31.6% over standard LoRA while remaining robust to rank choice.

Digital dentistry has made significant advances with the advent of deep learning. However, the majority of these deep learning-based dental image analysis models focus on very specific tasks such as tooth segmentation, tooth detection, cavity detection, and gingivitis classification. There is a lack of a specialized model that has holistic knowledge of teeth and can perform dental image analysis tasks based on that knowledge. Datasets of dental images with captions can help build such a model. To the best of our knowledge, existing dental image datasets with captions are few in number and limited in scope. In many of these datasets, the captions describe the entire mouth, while the images are limited to the anterior view. As a result, posterior teeth such as molars are not clearly visible, limiting the usefulness of the captions for training vision-language models. Additionally, the captions focus only on a specific disease (gingivitis) and do not provide a holistic assessment of each tooth. Moreover, tooth disease scores are typically assigned to individual teeth, and each tooth is treated as a separate entity in orthodontic procedures. Therefore, it is important to have captions for single-tooth images. As far as we know, no such dataset of single-tooth images with dental captions exists. In this work, we aim to bridge that gap by assessing the possibility of generating captions for dental images using Vision-Language Models (VLMs) and evaluating the extent and quality of those captions. Our findings suggest that guided prompts help VLMs generate meaningful captions. We show that the prompts generated by our framework are better anchored in describing the visual aspects of dental images. We selected RGB images as they have greater potential in consumer scenarios.

Linear autoencoders (LAEs) have gained increasing popularity in recommender systems due to their simplicity and strong empirical performance. Most LAE models, including the Emphasized Denoising Linear Autoencoder (EDLAE) introduced by (Steck, 2020), use quadratic loss during training. However, the original EDLAE only provides closed-form solutions for the hyperparameter choice $b = 0$, which limits its capacity. In this work, we generalize EDLAE objective into a Decoupled Expected Quadratic Loss (DEQL). We show that DEQL simplifies the process of deriving EDLAE solutions and reveals solutions in a broader hyperparameter range $b > 0$, which were not derived in Steck's original paper. Additionally, we propose an efficient algorithm based on Miller's matrix inverse theorem to ensure the computational tractability for the $b > 0$ case. Empirical results on benchmark datasets show that the $b > 0$ solutions provided by DEQL outperform the $b = 0$ EDLAE baseline, demonstrating that DEQL expands the solution space and enables the discovery of models with better testing performance.

Caption quality has emerged as a critical bottleneck in training high-quality text-to-image (T2I) and text-to-video (T2V) generative models. While visual language models (VLMs) are commonly deployed to generate captions from visual data, they suffer from hallucinations, poor compositional reasoning, and limited fine-grained understanding, resulting in misaligned image-caption pairs that degrade downstream model performance. This technical report introduces VIVECaption, a systematic two-sided approach to caption quality improvement. We first establish a comprehensive taxonomy of caption evaluation metrics, distinguishing between "universal" and "instance-grounded" metrics, with the ultimate goal of showcasing the use-cases and tradeoffs between different caption quality metrics. We then use this language to describe our two-sided approach to caption quality improvement: (1) a gold-standard dataset creation methodology using stratified sampling and (2) a model alignment strategy encompassing context alignment and parameter-level finetuning using SFT. We demonstrate our methodology on open-source models, focusing on structured caption formats that enable better parsing and downstream utilization. We ultimately show that using a finetuned character detection model in an image captioning pipeline significantly improves holistic image-caption alignment quality. Our work addresses the growing need for high-quality "vegan" training data in enterprise AI development, providing practical solutions for teams seeking to improve caption-image alignment without relying on potentially copyright-protected web-scraped content.

Many real-world walking scenarios contain obstacles and unsafe ground patches (e.g., slippery or cluttered areas), leaving a disconnected set of admissible footholds that can be modeled as stepping-stone-like regions. We propose an onboard, perceptive mixed-integer model predictive control framework that jointly plans foot placement and step duration using step-to-step Divergent Component of Motion (DCM) dynamics. Ego-centric depth images are fused into a probabilistic local heightmap, from which we extract a union of convex steppable regions. Region membership is enforced with binary variables in a mixed-integer quadratic program (MIQP). To keep the optimization tractable while certifying safety, we embed capturability bounds in the DCM space: a lateral one-step condition (preventing leg crossing) and a sagittal infinite-step bound that limits unstable growth. We further re-plan within the step by back-propagating the measured instantaneous DCM to update the initial DCM, improving robustness to model mismatch and external disturbances. We evaluate the approach in simulation on Digit on randomized stepping-stone fields, including external pushes. The planner generates terrain-aware, dynamically consistent footstep sequences with adaptive timing and millisecond-level solve times.

We are concerned with the challenge of reliably classifying and assessing intracranial aneurysms using deep learning without compromising clinical transparency. While traditional black-box models achieve high predictive accuracy, their lack of inherent interpretability remains a significant barrier to clinical adoption and regulatory approval. Explainability is paramount in medical modeling to ensure that AI-driven diagnoses align with established neurosurgical principles. Unlike traditional eXplainable AI (XAI) methods -- such as saliency maps, which often provide post-hoc, non-causal visual correlations -- Concept Bottleneck Models (CBMs) offer a robust alternative by constraining the model's internal logic to human-understandable clinical indices. In this article, we propose an end-to-end 3D Concept Bottleneck framework that maps high-dimensional neuroimaging features to a discrete set of morphological and hemodynamic concepts for aneurysm identification. We implemented this pipeline using a pre-trained 3D ResNet-34 backbone and a 3D DenseNet-121 to extract features from CTA volumes, which were subsequently processed through a soft bottleneck layer representing human-interpretable clinical concepts. The model was optimized using a joint-loss function to balance diagnostic focal loss and concept mean squared error (MSE), validated via stratified five-fold cross-validation. Our results demonstrate a peak task classification accuracy of 93.33% +/- 4.5% for the ResNet-34 architecture and 91.43% +/- 5.8% for the DenseNet-121 model. Furthermore, the implementation of 8-pass Test-Time Augmentation (TTA) yielded a robust mean accuracy of 88.31%, ensuring diagnostic stability during inference. By maintaining an accuracy-generalization gap of less than 0.04, this framework proves that high predictive performance can be achieved without sacrificing interpretability.

Visual Question Answering (VQA) is a core task for evaluating the capabilities of Vision-Language Models (VLMs). Existing VQA benchmarks primarily feature clear and unambiguous image-question pairs, whereas real-world scenarios often involve varying degrees of ambiguity that require nuanced reasoning and context-appropriate response strategies. Although recent studies have begun to address ambiguity in VQA, they lack (1) a systematic categorization of ambiguity levels and (2) datasets and models that support strategy-aware responses. In this paper, we introduce Ambiguous Visual Question Answering (AQuA), a fine-grained dataset that classifies ambiguous VQA instances into four levels according to the nature and degree of ambiguity, along with the optimal response strategy for each case. Our evaluation of diverse open-source and proprietary VLMs shows that most models fail to adapt their strategy to the ambiguity type, frequently producing overconfident answers rather than seeking clarification or acknowledging uncertainty. To address this challenge, we fine-tune VLMs on AQuA, enabling them to adaptively choose among multiple response strategies, such as directly answering, inferring intent from contextual cues, listing plausible alternatives, or requesting clarification. VLMs trained on AQuA achieve strategic response generation for ambiguous VQA, demonstrating the ability to recognize ambiguity, manage uncertainty, and respond with context-appropriate strategies, while outperforming both open-source and closed-source baselines.

Autonomous underwater robots are increasingly deployed for environmental monitoring, infrastructure inspection, subsea resource exploration, and long-horizon exploration. Yet, despite rapid advances in learning-based planning and control, reliable autonomy in real ocean environments remains fundamentally constrained by tightly coupled physical limits. Hydrodynamic uncertainty, partial observability, bandwidth-limited communication, and energy scarcity are not independent challenges; they interact within the closed perception-planning-control loop and often amplify one another over time. This Review develops a constraint-coupled perspective on underwater embodied intelligence, arguing that planning and control must be understood within tightly coupled sensing, communication, coordination, and resource constraints in real ocean environments. We synthesize recent progress in reinforcement learning, belief-aware planning, hybrid control, multi-robot coordination, and foundation-model integration through this embodied perspective. Across representative application domains, we show how environmental monitoring, inspection, exploration, and cooperative missions expose distinct stress profiles of cross-layer coupling. To unify these observations, we introduce a cross-layer failure taxonomy spanning epistemic, dynamic, and coordination breakdowns, and analyze how errors cascade across autonomy layers under uncertainty. Building on this structure, we outline research directions toward physics-grounded world models, certifiable learning-enabled control, communication-aware coordination, and deployment-aware system design. By internalizing constraint coupling rather than treating it as an external disturbance, underwater embodied intelligence may evolve from performance-driven adaptation toward resilient, scalable, and verifiable autonomy under real ocean conditions.

LLMs operating in dynamic real-world contexts often encounter knowledge that evolves continuously or emerges incrementally. To remain accurate and effective, models must adapt to newly arriving information on the fly. We introduce Online Adaptation to Continual Knowledge Streams(OAKS) to evaluate this capability, establishing a benchmark for online adaptation over streaming, continually updating knowledge. Specifically, the benchmark is structured as a sequence of fine-grained context chunks where facts change dynamically across time intervals. OAKS comprises two datasets: OAKS-BABI and OAKS-Novel, where individual facts evolve multiple times across context chunks. These datasets include dense annotations to measure whether models track changes accurately. Evaluating 14 models with varied inference approaches, we observe significant limitations in current methodologies. Both state-of-the-art models and agentic memory systems fail to adapt robustly on OAKS, demonstrating delays in state-tracking and susceptibility to distraction within streaming environments.

Legal teams increasingly use machine learning to triage large volumes of contractual evidence, but many models are opaque, non-deterministic, and difficult to align with frameworks such as HIPAA or NERC-CIP. We study a simple, reproducible alternative based on deterministic dual encoders and transparent fuzzy triage bands. We train a RoBERTa-base dual encoder with a 512-dimensional projection and cosine similarity on the ACORD benchmark for graded clause retrieval, then fine-tune it on a CUAD-derived binary compliance dataset. Across five random seeds (40-44) on a single NVIDIA A100 GPU, the model achieves ACORD-style retrieval performance of NDCG@5 0.38-0.42, NDCG@10 0.45-0.50, and 4-star Precision@5 about 0.37 on the test split. On CUAD-derived binary labels, it achieves AUC 0.98-0.99 and F1 0.22-0.30 depending on positive-class weighting, outperforming majority and random baselines in a highly imbalanced setting with a positive rate of about 0.6%. We then map scalar compliance scores into three regions: auto-noncompliant, auto-compliant, and human-review. Thresholds are tuned on validation data to maximize automatic decision coverage subject to an empirical error-rate constraint of at most 2% over auto-decided examples. The result is a seed-stable system summarized by a small number of scalar parameters. We argue that deterministic encoders, calibrated fuzzy bands, and explicit error constraints provide a practical middle ground between hand-crafted rules and opaque large language models, supporting explainable evidence triage, reproducible audit trails, and concrete mappings to legal review concepts.

In machine learning, the goal of multi-task learning (MTL) is to optimize multiple objectives together. Recent works, for example, Multiple Gradient Descent Algorithm (MGDA) and its variants, show promising results with dynamically adjusted weights for different tasks to mitigate conflicts that may potentially degrade the performance on certain tasks. Despite the empirical success of MGDA-type methods, one major limitation of such methods is their computational inefficiency, as they require access to all task gradients. In this paper we introduce MARIGOLD, a unified algorithmic framework for efficiently solving MTL problems. Our method reveals that multi-task gradient balancing methods have a hierarchical structure, in which the model training and the gradient balancing are coupled during the whole optimization process and can be viewed as a bi-level optimization problem. Moreover, we showcase that the bi-level problem can be solved efficiently by leveraging zeroth-order method. Extensive experiments on both public datasets and industrial-scale datasets demonstrate the efficiency and superiority of our method.

Retrieval-Augmented Generation (RAG) systems are increasingly evolving into agentic architectures where large language models autonomously coordinate multi-step reasoning, dynamic memory management, and iterative retrieval strategies. Despite rapid industrial adoption, current research lacks a systematic understanding of Agentic RAG as a sequential decision-making system, leading to highly fragmented architectures, inconsistent evaluation methodologies, and unresolved reliability risks. This Systematization of Knowledge (SoK) paper provides the first unified framework for understanding these autonomous systems. We formalize agentic retrieval-generation loops as finite-horizon partially observable Markov decision processes, explicitly modeling their control policies and state transitions. Building upon this formalization, we develop a comprehensive taxonomy and modular architectural decomposition that categorizes systems by their planning mechanisms, retrieval orchestration, memory paradigms, and tool-invocation behaviors. We further analyze the critical limitations of traditional static evaluation practices and identify severe systemic risks inherent to autonomous loops, including compounding hallucination propagation, memory poisoning, retrieval misalignment, and cascading tool-execution vulnerabilities. Finally, we outline key doctoral-scale research directions spanning stable adaptive retrieval, cost-aware orchestration, formal trajectory evaluation, and oversight mechanisms, providing a definitive roadmap for building reliable, controllable, and scalable agentic retrieval systems.

Quality Estimation (QE) is essential for assessing machine translation quality in reference-less settings, particularly for domain-specific and low-resource language scenarios. In this paper, we investigate sentence-level QE for English to Indic machine translation across four domains (Healthcare, Legal, Tourism, and General) and five language pairs. We systematically compare zero-shot, few-shot, and guideline-anchored prompting across selected closed-weight and open-weight LLMs. Findings indicate that while closed-weight models achieve strong performance via prompting alone, prompt-only approaches remain fragile for open-weight models, especially in high-risk domains. To address this, we adopt ALOPE, a framework for LLM-based QE that uses Low-Rank Adaptation with regression heads attached to selected intermediate Transformer layers. We also extend ALOPE with recently proposed Low-Rank Multiplicative Adaptation (LoRMA). Our results show that intermediate-layer adaptation consistently improves QE performance, with gains in semantically complex domains, indicating a path toward more robust QE in practical scenarios. We release code and domain-specific QE datasets publicly to support further research.

Reconfigurable Intelligent Surfaces promise to transform wireless environments, yet practical deployment is hindered by the prohibitive overhead of Channel State Information (CSI) estimation and the dimensionality explosion inherent in centralized optimization. This paper proposes a Hierarchical Multi-Agent Reinforcement Learning (HMARL) framework for the control of mechanically reconfigurable reflective surfaces in millimeter-wave (mmWave) systems. We introduce a "CSI-free" paradigm that substitutes pilot-based channel estimation with readily available user localization data. To manage the massive combinatorial action space, the proposed architecture utilizes Multi-Agent Proximal Policy Optimization (MAPPO) under a Centralized Training with Decentralized Execution (CTDE) paradigm. The proposed architecture decomposes the control problem into two abstraction levels: a high-level controller for user-to-reflector allocation and decentralized low-level controllers for low-level focal point optimization. Comprehensive ray-tracing evaluations demonstrate that the framework achieves 2.81-7.94 dB RSSI improvements over centralized baselines, with the performance advantage widening as system complexity increases. Scalability analysis reveals that the system maintains sustained efficiency, exhibiting minimal per-user performance degradation and stable total power utilization even when user density doubles. Furthermore, robustness validation confirms the framework's viability across varying reflector aperture sizes (45-99 tiles) and demonstrates graceful performance degradation under localization errors up to 0.5 m. By eliminating CSI overhead while maintaining high-fidelity beam-focusing, this work establishes HMARL as a practical solution for intelligent mmWave environments.