Checklists have emerged as a popular approach for interpretable and fine-grained evaluation, particularly with LLM-as-a-Judge. Beyond evaluation, these structured criteria can serve as signals for model alignment, reinforcement learning, and self-correction. To support these use cases, we present AutoChecklist, an open-source library that unifies checklist-based evaluation into composable pipelines. At its core is a taxonomy of five checklist generation abstractions, each encoding a distinct strategy for deriving evaluation criteria. A modular Generator $\rightarrow$ Refiner $\rightarrow$ Scorer pipeline connects any generator with a unified scorer, and new configurations can be registered via prompt templates alone. The library ships with ten built-in pipelines implementing published approaches and supports multiple LLM providers (OpenAI, OpenRouter, vLLM). Beyond the Python API, the library includes a CLI for off-the-shelf evaluation and a web interface for interactive exploration. Validation experiments confirm that these checklist methods significantly align with human preferences and quality ratings, and a case study on ICLR peer review rebuttals demonstrates flexible domain adaptation. AutoChecklist is publicly available at https://github.com/ChicagoHAI/AutoChecklist.

Safety alignment is critical for deploying large language models (LLMs) in real-world applications, yet most existing approaches rely on large human-annotated datasets and static red-teaming benchmarks that are costly, difficult to scale, and slow to adapt to evolving model behaviors. Moreover, overly conservative safety mechanisms can reduce model usefulness by rejecting sensitive but legitimate queries. We introduce Self-MOA (Self Multi-Objective Alignment), a fully automated framework for aligning small language models using weak supervision from automated evaluator models. Self-MOA operates as a closed loop that dynamically generates model-specific red team prompts, constructs preference data from model-generated responses, and aligns models via multi-objective preference optimization to jointly optimize for safety and helpfulness. Across multiple small language models and safety benchmarks, Self-MOA achieves a 12.41\% improvement in safety while preserving helpfulness, using as little as 11 times less training data than human-supervised alignment baselines. These results demonstrate that adaptive, automated alignment can reduce the dependence on static, human-curated safety pipelines in resource-constrained settings.

A matching platform is a system that matches different types of participants, such as companies and job-seekers. In such a platform, merely maximizing the number of matches can result in matches being concentrated on highly popular participants, which may increase dissatisfaction among other participants, such as companies, and ultimately lead to their churn, reducing the platform's profit opportunities. To address this issue, we propose a novel online learning problem, Combinatorial Allocation Bandits (CAB), which incorporates the notion of *arm satisfaction*. In CAB, at each round $t=1,\dots,T$, the learner observes $K$ feature vectors corresponding to $K$ arms for each of $N$ users, assigns each user to an arm, and then observes feedback following a generalized linear model (GLM). Unlike prior work, the learner's objective is not to maximize the number of positive feedback, but rather to maximize the arm satisfaction. For CAB, we provide an upper confidence bound algorithm that achieves an approximate regret upper bound, which matches the existing lower bound for the special case. Furthermore, we propose a TS algorithm and provide an approximate regret upper bound. Finally, we conduct experiments on synthetic data to demonstrate the effectiveness of the proposed algorithms compared to other methods.

Efficient path following for mobile manipulators is often hindered by high-dimensional configuration spaces and kinematic constraints. This paper presents a robust two-stage configuration planning framework that decouples the 8-DoF planning problem into a tractable 2-DoF base optimization under a yaw-fixed base planning assumption. In the first stage, the proposed approach utilizes IRM to discretize the task-space path into a multi-layer graph, where an initial feasible path is extracted via a Dijkstra-based dynamic programming approach to ensure computational efficiency and global optimality within the discretized graph. In the second stage, to overcome discrete search quantization, feasible base regions are transformed into convex hulls, enabling subsequent continuous refinement via the L-BFGS algorithm to maximize trajectory smoothness while strictly enforcing reachability constraints. Simulation results demonstrate the theoretical precision of the proposed method by achieving sub-millimeter kinematic accuracy in simulation, and physical experiments on an omnidirectional mobile manipulator further validate the framework's robustness and practical applicability.

Recent advances in image synthesis have been propelled by powerful generative models, such as Masked Generative Transformers (MaskGIT), autoregressive models, diffusion models, and rectified flow models. A common principle behind their success is the decomposition of synthesis into multiple steps. However, this introduces a proliferation of step-specific parameters (e.g., noise level or temperature at each step). Existing approaches typically rely on manually-designed rules to manage this complexity, demanding expert knowledge and trial-and-error. Furthermore, these static schedules lack the flexibility to adapt to the unique characteristics of each sample, yielding sub-optimal performance. To address this issue, we present AdaGen, a general, learnable, and sample-adaptive framework for scheduling the iterative generation process. Specifically, we formulate the scheduling problem as a Markov Decision Process, where a lightweight policy network determines suitable parameters given the current generation state, and can be trained through reinforcement learning. Importantly, we demonstrate that simple reward designs, such as FID or pre-trained reward models, can be easily hacked and may not reliably guarantee the desired quality or diversity of generated samples. Therefore, we propose an adversarial reward design to guide the training of the policy networks. Finally, we introduce an inference-time refinement strategy and a controllable fidelity-diversity trade-off mechanism to further enhance the performance and flexibility of AdaGen. Comprehensive experiments on four generative paradigms validate the superiority of AdaGen. For example, AdaGen achieves better performance on DiT-XL with 3 times lower inference cost and improves the FID of VAR from 1.92 to 1.59 with negligible computational overhead.

In predictive modeling for low-resource audio classification, extracting high-accuracy and interpretable attributes is critical. Particularly in high-reliability applications, interpretable audio attributes are indispensable. While human-driven attribute discovery is effective, its low throughput becomes a bottleneck. We propose a method for adaptively discovering interpretable audio attributes using Multimodal Large Language Models (MLLMs). By replacing humans in the AdaFlock framework with MLLMs, our method achieves significantly faster attribute discovery. Our method dynamically identifies salient acoustic characteristics via prompting and constructs an attribute-based ensemble classifier. Experimental results across various audio tasks demonstrate that our method outperforms direct MLLM prediction in the majority of evaluated cases. The entire training completes within 11 minutes, proving it a practical, adaptive solution that surpasses conventional human-reliant approaches.

Deploying visuomotor robots at scale is challenging due to the potential for anomalous failures to degrade performance, cause damage, or endanger human life. Bimanual manipulators are no exception; these robots have vast state spaces comprised of high-dimensional images and proprioceptive signals. Explicitly defining failure modes within such state spaces is infeasible. In this work, we overcome these challenges by training a probabilistic, history informed, world model within the compressed latent space of a pretrained vision foundation model (NVIDIA's Cosmos Tokenizer). The model outputs uncertainty estimates alongside its predictions that serve as non-conformity scores within a conformal prediction framework. We use these scores to develop a runtime monitor, correlating periods of high uncertainty with anomalous failures. To test these methods, we use the simulated Push-T environment and the Bimanual Cable Manipulation dataset, the latter of which we introduce in this work. This new dataset features trajectories with multiple synchronized camera views, proprioceptive signals, and annotated failures from a challenging data center maintenance task. We benchmark our methods against baselines from the anomaly detection and out-of-distribution detection literature, and show that our approach considerably outperforms statistical techniques. Furthermore, we show that our approach requires approximately one twentieth of the trainable parameters as the next-best learning-based approach, yet outperforms it by 3.8% in terms of failure detection rate, paving the way toward safely deploying manipulator robots in real-world environments where reliability is non-negotiable.

Spatial reasoning from monocular images is essential for autonomous driving, yet current Vision-Language Models (VLMs) still struggle with fine-grained geometric perception, particularly under large scale variation and ambiguous object appearance. We propose a simple yet effective perception-aware multimodal reasoning framework that equips VLMs with explicit object-centric grounding ability. Instead of relying on textual bounding-box outputs, each referred object is represented using all Visual Reference Tokens (VRTs) within its spatial extent, enabling visual evidence and textual reasoning to be processed jointly in a unified token space. To further strengthen cross-modal interaction, we construct a Multimodal Chain-of-Thought (MM-CoT) dataset that injects aligned visual and textual reasoning signals. A deterministic ordering strategy is introduced to make supervision over inherently unordered VRT sets fully compatible with the VLM's autoregressive next-token prediction. With only standard supervised fine-tuning, our method achieves substantial improvements on the SURDS benchmark, outperforming previous approaches - including those using RL-based post-training - by a large margin across both single-object and multi-object tasks. These results demonstrate that accurate perception and multimodal reasoning are mutually reinforcing, and together form the key to robust spatial understanding in challenging monocular driving scenarios.

Image-based shape retrieval (IBSR) aims to retrieve 3D models from a database given a query image, hence addressing a classical task in computer vision, computer graphics, and robotics. Recent approaches typically rely on bridging the domain gap between 2D images and 3D shapes based on the use of multi-view renderings as well as task-specific metric learning to embed shapes and images into a common latent space. In contrast, we address IBSR through large-scale multi-modal pretraining and show that explicit view-based supervision is not required. Inspired by pre-aligned image--point-cloud encoders from ULIP and OpenShape that have been used for tasks such as 3D shape classification, we propose the use of pre-aligned image and shape encoders for zero-shot and standard IBSR by embedding images and point clouds into a shared representation space and performing retrieval via similarity search over compact single-embedding shape descriptors. This formulation allows skipping view synthesis and naturally enables zero-shot and cross-domain retrieval without retraining on the target database. We evaluate pre-aligned encoders in both zero-shot and supervised IBSR settings and additionally introduce a multi-modal hard contrastive loss (HCL) to further increase retrieval performance. Our evaluation demonstrates state-of-the-art performance, outperforming related methods on $Acc_{Top1}$ and $Acc_{Top10}$ for shape retrieval across multiple datasets, with best results observed for OpenShape combined with Point-BERT. Furthermore, training on our proposed multi-modal HCL yields dataset-dependent gains in standard instance retrieval tasks on shape-centric data, underscoring the value of pretraining and hard contrastive learning for 3D shape retrieval. The code will be made available via the project website.

Controllable diffusion generation often relies on various heuristics that are seemingly disconnected without a unified understanding. We bridge this gap with Diffusion Controller (DiffCon), a unified control-theoretic view that casts reverse diffusion sampling as state-only stochastic control within (generalized) linearly-solvable Markov Decision Processes (LS-MDPs). Under this framework, control acts by reweighting the pretrained reverse-time transition kernels, balancing terminal objectives against an $f$-divergence cost. From the resulting optimality conditions, we derive practical reinforcement learning methods for diffusion fine-tuning: (i) f-divergence-regularized policy-gradient updates, including a PPO-style rule, and (ii) a regularizer-determined reward-weighted regression objective with a minimizer-preservation guarantee under the Kullback-Leibler (KL) divergence. The LS-MDP framework further implies a principled model form: the optimal score decomposes into a fixed pretrained baseline plus a lightweight control correction, motivating a side-network parameterization conditioned on exposed intermediate denoising outputs, enabling effective gray-box adaptation with a frozen backbone. Experiments on Stable Diffusion v1.4 across supervised and reward-driven finetuning show consistent gains in preference-alignment win rates and improved quality-efficiency trade-offs versus gray-box baselines and even the parameter-efficient white-box adapter LoRA.

Soft robots are compliant but often cannot support loads or hold their shape, while rigid robots provide structural strength but are less adaptable. Existing variable-stiffness systems usually operate at the scale of whole segments or patches, which limits precise control over stiffness distribution and virtual joint placement. This paper presents the Variable Stiffness Lattice Skin (VSL-Skin), the first system to enable individually addressable voxel-level morphological control with centimeter-scale precision. The system provides three main capabilities: nearly two orders of magnitude stiffness modulation across axial (15-1200 N/mm), shear (45-850 N/mm), bending (8*10^2 - 3*10^4 N/deg), and torsional modes with centimeter-scale spatial control; the first demonstrated 30% axial compression in phase-change systems while maintaining structural integrity; and autonomous component-level self-repair through thermal cycling, which eliminates fatigue accumulation and enables programmable sacrificial joints for predictable failure management. Selective voxel activation creates six canonical virtual joint types with programmable compliance while preserving structural integrity in non-activated regions. The platform incorporates closed-form design models and finite element analysis for predictive synthesis of stiffness patterns and joint placement. Experimental validation demonstrates 30% axial contraction, thermal switching in 75-second cycles, and cut-to-fit integration that preserves addressability after trimming. The row-column architecture enables platform-agnostic deployment across diverse robotic systems without specialized infrastructure. This framework establishes morphological intelligence as an engineerable system property and advances autonomous reconfigurable robotics.

We present the first large-scale, cross-domain evaluation of document chunking strategies for dense retrieval, addressing a critical but underexplored aspect of retrieval-augmented systems. In our study, 36 segmentation methods spanning fixed-size, semantic, structure-aware, hierarchical, adaptive, and LLM-assisted approaches are benchmarked across six diverse knowledge domains using five different embedding models. Retrieval performance is assessed using graded relevance scores from a state-of-the-art LLM evaluator, with Normalised DCG@5 as the primary metric (complemented by Hit@5 and MRR). Our experiments show that content-aware chunking significantly improves retrieval effectiveness over naive fixed-length splitting. The top-performing strategy, Paragraph Group Chunking, achieved the highest overall accuracy (mean nDCG@5~0.459) and substantially better top-rank hit rates (Precision@1~24%, Hit@5~59%). In contrast, simple fixed-size character chunking as baselines performed poorly (nDCG@5 < 0.244, Precision@1~2-3%). We observe pronounced domain-specific differences: dynamic token sizing is strongest in biology, physics and health, while paragraph grouping is strongest in legal and maths. Larger embedding models yield higher absolute scores but remain sensitive to suboptimal segmentation, indicating that better chunking and large embeddings provide complementary benefits. In addition to accuracy gains, we quantify the efficiency trade-offs of advanced chunking. Producing more, smaller chunks can increase index size and latency. Consequently, we identify methods (like dynamic chunking) that approach an optimal balance of effectiveness and efficiency. These findings establish chunking as a vital lever for improving retrieval performance and reliability.

We present Elenchus, a dialogue system for knowledge base construction grounded in inferentialist semantics, where knowledge engineering is re-conceived as explicitation rather than extraction from expert testimony or textual content. A human expert develops a bilateral position (commitments and denials) about a topic through prover-skeptic dialogue with a large language model (LLM) opponent. The LLM proposes tensions (claims that parts of the position are jointly incoherent) which the expert resolves by retraction, refinement, or contestation. The LLM thus serves as a defeasible derivability oracle whose unreliability is structurally contained by the expert's authority. Our main technical contribution is a mapping from Elenchus dialectical states to material bases in Hlobil and Brandom's NonMonotonic MultiSuccedent (NMMS) logic, satisfying Containment and enabling the elaboration of logical vocabulary that makes explicit the inferential relationships negotiated in the dialectic. We demonstrate the approach on the W3C PROV-O provenance ontology, where a single dialogue session elicits and structures design tensions that a domain expert can articulate, corresponding to decisions documented in a retrospective analysis of the ontology's design. Using pyNMMS, an automated NMMS reasoner, we verify that the structural properties of the resulting material base (nontransitivity, nonmonotonicity, and independence) correspond to specific PROV design rationales, demonstrating end-to-end integration from dialogue through formal reasoning.

Video Temporal Grounding (VTG) aims to localize the video segment that corresponds to a natural language query, which requires a comprehensive understanding of complex temporal dynamics. Existing Vision-LMMs typically perceive temporal dynamics via positional encoding, text-based timestamps, or visual frame numbering. However, these approaches exhibit notable limitations: assigning each frame a text-based timestamp token introduces additional computational overhead and leads to sparsity in visual attention, positional encoding struggles to capture absolute temporal information, and visual frame numbering often compromises spatial detail. To address these issues, we propose Temporal to Spatial Gridification (T2SGrid), a novel framework that reformulates video temporal understanding as a spatial understanding task. The core idea of T2SGrid is to process video content in clips rather than individual frames. we employ a overlapping sliding windows mechanism to segment the video into temporal clips. Within each window, frames are arranged chronologically in a row-major order into a composite grid image, effectively transforming temporal sequences into structured 2D layouts. The gridification not only encodes temporal information but also enhances local attention within each grid. Furthermore, T2SGrid enables the use of composite text timestamps to establish global temporal awareness. Experiments on standard VTG benchmarks demonstrate that T2SGrid achieves superior performance.

Reconstructing surgical scenes from monocular endoscopic video is critical for advancing robotic-assisted surgery. However, the application of state-of-the-art general-purpose reconstruction models is constrained by two key challenges: the lack of supervised training data and performance degradation over long video sequences. To overcome these limitations, we propose SurgCUT3R, a systematic framework that adapts unified 3D reconstruction models to the surgical domain. Our contributions are threefold. First, we develop a data generation pipeline that exploits public stereo surgical datasets to produce large-scale, metric-scale pseudo-ground-truth depth maps, effectively bridging the data gap. Second, we propose a hybrid supervision strategy that couples our pseudo-ground-truth with geometric self-correction to enhance robustness against inherent data imperfections. Third, we introduce a hierarchical inference framework that employs two specialized models to effectively mitigate accumulated pose drift over long surgical videos: one for global stability and one for local accuracy. Experiments on the SCARED and StereoMIS datasets demonstrate that our method achieves a competitive balance between accuracy and efficiency, delivering near state-of-the-art but substantially faster pose estimation and offering a practical and effective solution for robust reconstruction in surgical environments. Project page: https://chumo-xu.github.io/SurgCUT3R-ICRA26/.

Quadruped locomotion provides a natural setting for understanding when model-free learning can outperform model-based control design, by exploiting data patterns to bypass the difficulty of optimizing over discrete contacts and the combinatorial explosion of mode changes. We give a principled analysis of why imitation learning with quadrupeds can be inherently effective in a small data regime, based on the structure of its limit cycles, Poincaré return maps, and local numerical properties of neural networks. The understanding motivates a new imitation learning method that regulates the alignment between variations in a latent space and those over the output actions. Hardware experiments confirm that a few seconds of demonstration is sufficient to train various locomotion policies from scratch entirely offline with reasonable robustness.

Accurate chart comprehension represents a critical challenge in advancing multimodal learning systems, as extensive information is compressed into structured visual representations. However, existing vision-language models (VLMs) frequently struggle to generalize on unseen charts because it requires abstract, symbolic, and quantitative reasoning over structured visual representations. In this work, we introduce Chart-RL, an effective reinforcement learning (RL) method that employs mathematically verifiable rewards to enhance chart question answering in VLMs. Our experiments demonstrate that Chart-RL consistently outperforms supervised fine-tuning (SFT) across different chart understanding benchmarks, achieving relative improvements of 16.7% on MutlChartQA, and 11.5% on ChartInsights. We conduct robustness analysis, where Chart-RL achieves enhanced performance in 18 of 25 perturbed chart categories, demonstrating strong consistency and reasoning capability across visual variations. Furthermore, we demonstrate that task difficulty and inherent complexity are more critical than data quantity in RL training. For instance, Chart-RL trained on merely 10 complex chart-query examples significantly outperforms models trained on over 6,000 simple examples. Additionally, training on challenging reasoning tasks not only improves in-domain generalization relative to simpler tasks, but also facilitate strong transfer to out-of-domain visual mathematical problems.

Purpose: Incomplete dissection is a common cause of persistent disease and revision endoscopic sinus surgery (ESS) in chronic rhinosinusitis. Current image-guided surgery systems typically reference static preoperative CT (pCT), and do not model evolving resection boundaries. We present Virtual Intraoperative CT (viCT), a method for sequentially updating pCT throughout ESS using intraoperative 3D reconstructions from monocular endoscopic video to enable visualization of evolving anatomy in CT format. Methods: Monocular endoscopic video is processed using a depth-supervised NeRF framework with virtual stereo synthesis to generate metrically scaled 3D reconstructions at multiple surgical intervals. Reconstructions undergo rigid, landmark-based registration in 3D Slicer guided by anatomical correspondences, and are then voxelized into the pCT grid. viCT volumes were generated using a ray-based occupancy comparison between pCT and reconstruction to delete outdated voxels and remap preserved anatomy and updated boundaries. Performance is evaluated in a cadaveric feasibility study of four specimens across four ESS stages using volumetric overlap (DSC, Jaccard) and surface metrics (HD95, Chamfer, MSD, RMSD), and qualitative comparisons to ground-truth CT. Results: viCT updates show agreement with ground-truth anatomy across surgical stages, with submillimeter mean surface errors. Dice Similarity Coefficient (DSC) = 0.88 +/- 0.05 and Jaccard Index = 0.79 +/- 0.07, and Hausdorff Distance 95% (HD95) = 0.69 +/- 0.28 mm, Chamfer Distance = 0.09 +/- 0.05 mm, Mean Surface Distance (MSD) = 0.11 +/- 0.05 mm, and Root Mean Square Distance (RMSD) = 0.32 +/- 0.10 mm. Conclusion: viCT enables CT-format anatomic updating in an ESS setting without ancillary hardware. Future work will focus on fully automating registration, validation in live cases, and optimizing runtime for real-time deployment.

Collaborative aerial transportation of tethered payloads is fundamentally limited by space, power, and weight constraints. Conventional approaches rely on static equilibrium conditions, where each vehicle tilts to generate the forces that ensure they maintain a formation geometry that avoids aerodynamic interactions and collision. This horizontal thrust component represents a significant energy penalty compared to the ideal case in which each vehicle produces purely vertical thrust to lift the payload. Operating in tighter tether configurations can minimize this effect, but at the cost of either having to fly the vehicles in closer proximity, which risks collision, or significantly increasing the length of the tether, which increases complexity and reduces potential use-cases. We propose operating the tether-suspended flying system at a rotating equilibrium. By maintaining steady circular motion, centrifugal forces provide the necessary horizontal tether tension, allowing each quadrotor to generate purely vertical thrust and thus reducing the total force (and power) required compared to an equilibrium where the thrusts are not vertical. It also allows for a wider range of tether configurations to be used without sacrificing efficiency. Results demonstrate that rotating equilibria can reduce power consumption relative to static lifting by up to 20%, making collaborative aerial solutions more practically relevant.

This tutorial provides a critical review of the practical application of Control Barrier Functions (CBFs) in robotic safety. While the theoretical foundations of CBFs are well-established, I identify a recurring gap between the mathematical assumption of a safe controller's existence and its constructive realization in systems with input constraints. I highlight the distinction between candidate and valid CBFs by analyzing the interplay of system dynamics, actuation limits, and class-K functions. I further show that some purported demonstrations of safe robot policies or controllers are limited to passively safe systems, such as single integrators or kinematic manipulators, where safety is already inherited from the underlying physics and even naive geometric hard constraints suffice to prevent collisions. By revisiting simple low-dimensional examples, I show when CBF formulations provide valid safety guarantees and when they fail due to common misuses. I then provide practical guidelines for constructing realizable safety arguments for systems without such passive safety. The goal of this tutorial is to bridge the gap between theoretical guarantees and actual implementation, supported by an open-source interactive web demonstration that visualizes these concepts intuitively.

As graphs scale to billions of nodes and edges, graph Machine Learning workloads are constrained by the cost of multi-hop traversals over exponentially growing neighborhoods. While various system-level and algorithmic optimizations have been proposed to accelerate Graph Neural Network (GNN) pipelines, data management and movement remain the primary bottlenecks at scale. In this paper, we explore whether graph sparsification, a well-established technique that reduces edges to create sparser neighborhoods, can serve as a lightweight pre-processing step to address these bottlenecks while preserving accuracy on node classification tasks. We develop an extensible experimental framework that enables systematic evaluation of how different sparsification methods affect the performance and accuracy of GNN models. We conduct the first comprehensive study of GNN training and inference on sparsified graphs, revealing several key findings. First, sparsification often preserves or even improves predictive performance. As an example, random sparsification raises the accuracy of the GAT model by 6.8% on the PubMed graph. Second, benefits increase with scale, substantially accelerating both training and inference. Our results show that the K-Neighbor sparsifier improves model serving performance on the Products graph by 11.7x with only a 0.7% accuracy drop. Importantly, we find that the computational overhead of sparsification is quickly amortized, making it practical for very large graphs.

Signal Temporal Logic (STL) is expressive formal language that specifies spatio-temporal requirements in robotics. Its quantitative robustness semantics can be easily integrated with optimization-based control frameworks. However, STL specifications may become conflicting in real-world applications, where safety rules, traffic regulations, and task objectives can be cannot be satisfied together. In these situations, traditional STL-constrained Model Predictive Control (MPC) becomes infeasible and default to conservative behaviors such as freezing, which can largely increase risks in safety-critical scenarios. In this paper, we proposes a unified two-stage framework that first restores feasibility via minimal relaxation, then refine the feasible solution by formulating it as a value-aware multi-objective optimization problem. Using $\varepsilon$-constraint method, we approximate the Pareto front of the multi-objective optimization, which allows analysis of tradeoffs among competing objectives and counterfactual analysis of alternative actions. We demonstrate that the proposed approach avoids deadlock under conflicting STL specifications and enables interpretable decision-making in safety-critical applications by conducting a case study in autonomous driving.

Many distributional quantities in reinforcement learning are intrinsically joint across actions, including distributions of gaps and probabilities of superiority. However, the classical Markov decision process (MDP) formalism specifies only marginal laws and leaves the joint law of counterfactual one-step outcomes across multiple possible actions at a state unspecified. We study coupled-dynamics environments with a multi-action generative interface which can sample counterfactual one-step outcomes for multiple actions under shared exogenous randomness. We propose joint MDPs (JMDPs) as a formalism for such environments by augmenting an MDP with a multi-action sample transition model which specifies a coupling of one-step counterfactual outcomes, while preserving standard MDP interaction as marginal observations. We adopt and formalize a one-step coupling regime where dependence across actions is confined to immediate counterfactual outcomes at the queried state. In this regime, we derive Bellman operators for $n$th-order return moments, providing dynamic programming and incremental algorithms with convergence guarantees.

Recent advances have made long-form report-generating systems widely available. This has prompted evaluation frameworks that use LLM-as-judge protocols and claim verification, along with meta-evaluation frameworks that seek to validate these methods. Many of the meta-evaluations estimate an evaluation quality's by comparing its assessments against human pairwise preferences. Prior work, however, suggests that human pairwise preference may be overly simplistic and can fail to capture nuances of expert expectations. We conduct a case study in meta-evaluation for long-form QA benchmarks using ScholarQA-CS2, a benchmark designed for assessing retrieval-augmented deep-research QA in the scientific domain. We comprehensively validate the benchmark through human pairwise preference judgments, then critically examine the strengths, weaknesses, and confounders of this approach. We show that pairwise preference rankings are best suited for system-level evaluation, while explicit metric-wise annotations and expert annotators are critical for reliable metric-level assessment, with subjectivity remaining a key challenge. Based on our findings, we offer practical guidelines for designing future meta-evaluations that better align evaluation methods, annotator expertise, and reporting practices. By surfacing these methodological challenges, we aim to advance evaluation standards for deep-research systems.

In this work, we study the mechanical behavior of solids with microstructure using the framework of Cosserat elasticity with a single unit director. This formulation captures the coupling between deformation and orientational fields that arises in many structured materials. To compute equilibrium configurations of such media, we develop two complementary computational approaches: a finite element formulation based on variational principles and a neural network-based solver that directly minimizes the total potential energy. The neural architecture is constructed to respect the fundamental kinematic structure of the theory. In particular, it enforces frame invariance of the energy, satisfies the unit-length constraint on the director field, and represents deformation and director fields through separate networks to preserve their kinematic independence in the variational setting. Beyond satisfying balance laws, however, physically admissible solutions must also correspond to stable energy minimizers. To assess this requirement, we derive the quasiconvexity condition, rank-one convexity condition, and the Legendre-Hadamard inequalities for the Cosserat model and formulate them in a manner suitable for evaluating neural network predictions. These necessary stability conditions provide a physics-based validation framework: network outputs that violate these necessary conditions cannot correspond to stable energy minimizers and can therefore be rejected. In this way, we integrate classical variational stability theory with modern machine-learning solvers, establishing a computational workflow in which equilibrium solutions are not only learned but also assessed for energetic consistency.

Mixture-of-experts (MoE) is a common approach for increasing parameter capacity, but applying MoE to state space model (SSM) token mixers can multiply the cost of the recurrent state update. We study how to introduce expert specialization into selective SSMs while preserving computational efficiency. We show that MoE--SSM can refer to two designs: (1) MoE over separated SSMs, which maintains multiple state trajectories and thus scales compute with the number of experts; and (2) MoE-parameterized SSM, which mixes experts in parameter space, maintains a single state trajectory, and evaluates the recurrence once. Our method, Switch Mamba (Swimba), follows the second design by routing over expert-produced SSM streams. Theoretically, we establish well-definedness and stability for MoE-parameterized SSMs and characterize the relationship between the two designs. Empirically, we evaluate Swimba on standard benchmark tasks and measure real-time throughput and latency. Under matched FLOPs, Swimba achieves slightly better average performance than the baseline, with a small slowdown in real-time latency and throughput. Overall, these results suggest that parameter-space MoE can increase SSM capacity while keeping the dominant recurrence cost fixed.

Diagnostic grading of prostate cancer (PCa) relies on the examination of 2D histology sections. However, the limited sampling of specimens afforded by 2D histopathology, and ambiguities when viewing 2D cross-sections, can lead to suboptimal treatment decisions. Recent studies have shown that 3D histomorphometric analysis of glands and nuclei can improve PCa risk assessment compared to analogous 2D features. Here, we expand on these efforts by developing an analytical pipeline to extract 3D features related to perineural invasion (PNI) and lymphovascular invasion (LVI), which correlate with poor prognosis for a variety of cancers. A 3D segmentation model (nnU-Net) was trained to segment nerves and vessels in 3D datasets of archived prostatectomy specimens that were optically cleared, labeled with a fluorescent analog of H&E, and imaged with open-top light-sheet (OTLS) microscopy. PNI- and LVI-related features, including metrics describing cancer-nerve and cancer-vessel proximity, were then extracted based on the 3D nerve/vessel segmentation masks in conjunction with 3D masks of cancer-enriched regions. As a preliminary exploration of the prognostic value of these features, we trained a supervised machine learning classifier to predict 5-year biochemical recurrence (BCR) outcomes, finding that 3D PNI-related features are moderately prognostic and outperform 2D PNI-related features (AUC = 0.71 vs. 0.52). Source code is available at https://github.com/sarahrahsl/SegCIA.git.

Dataset distillation often prioritizes global semantic proximity when creating small surrogate datasets for original large-scale ones. However, object semantics are inherently hierarchical. For example, the position and appearance of a bird's eyes are constrained by the outline of its head. Global proximity alone fails to capture how object-relevant structures at different levels support recognition. In this work, we investigate the contributions of hierarchical semantics to effective distilled data. We leverage the vision autoregressive (VAR) model whose coarse-to-fine generation mirrors this hierarchy and propose HIERAMP to amplify semantics at different levels. At each VAR scale, we inject class tokens that dynamically identify salient regions and use their induced maps to guide amplification at that scale. This adds only marginal inference cost while steering synthesis toward discriminative parts and structures. Empirically, we find that semantic amplification leads to more diverse token choices in constructing coarse-scale object layouts. Conversely, at fine scales, the amplification concentrates token usage, increasing focus on object-related details. Across popular dataset distillation benchmarks, HIERAMP consistently improves validation performance without explicitly optimizing global proximity, demonstrating the importance of semantic amplification for effective dataset distillation.

Indoor traversability segmentation aims to identify safe, navigable free space for autonomous agents, which is critical for robotic navigation. Pure vision-based models often fail to detect thin obstacles, such as chair legs, which can pose serious safety risks. We propose a multi-modal segmentation framework that leverages RGB images and sparse 1D laser depth information to capture geometric interactions and improve the detection of challenging obstacles. To reduce the reliance on large labeled datasets, we adopt the few-shot segmentation (FSS) paradigm, enabling the model to generalize from limited annotated examples. Traditional FSS methods focus solely on positive prototypes, often leading to overfitting to the support set and poor generalization. To address this, we introduce a negative contrastive learning (NCL) branch that leverages negative prototypes (obstacles) to refine free-space predictions. Additionally, we design a two-stage attention depth module to align 1D depth vectors with RGB images both horizontally and vertically. Extensive experiments on our custom-collected indoor RGB-D traversability dataset demonstrate that our method outperforms state-of-the-art FSS and RGB-D segmentation baselines, achieving up to 9\% higher mIoU under both 1-shot and 5-shot settings. These results highlight the effectiveness of leveraging negative prototypes and sparse depth for robust and efficient traversability segmentation.

Targets in remote sensing images are usually small, weakly textured, and easily disturbed by complex backgrounds, challenging high-precision detection with general algorithms. Building on our earlier ESM-YOLO, this work presents ESM-YOLO+ as a lightweight visible infrared fusion network. To enhance detection, ESM-YOLO+ includes two key innovations. (1) A Mask-Enhanced Attention Fusion (MEAF) module fuses features at the pixel level via learnable spatial masks and spatial attention, effectively aligning RGB and infrared features, enhancing small-target representation, and alleviating cross-modal misalignment and scale heterogeneity. (2) Training-time Structural Representation (SR) enhancement provides auxiliary supervision to preserve fine-grained spatial structures during training, boosting feature discriminability without extra inference cost. Extensive experiments on the VEDAI and DroneVehicle datasets validate ESM-YOLO+'s superiority. The model achieves 84.71\% mAP on VEDAI and 74.0\% mAP on DroneVehicle, while greatly reducing model complexity, with 93.6\% fewer parameters and 68.0\% lower GFLOPs than the baseline. These results confirm that ESM-YOLO+ integrates strong performance with practicality for real-time deployment, providing an effective solution for high-performance small-target detection in complex remote sensing scenes.