Active Surveillance (AS) is a treatment option for managing low and intermediate-risk prostate cancer (PCa), aiming to avoid overtreatment while monitoring disease progression through serial MRI and clinical follow-up. Accurate prostate segmentation is an important preliminary step for automating this process, enabling automated detection and diagnosis of PCa. However, existing deep-learning segmentation models are often trained on single-time-point and expertly annotated datasets, making them unsuitable for longitudinal AS analysis, where multiple time points and a scarcity of expert labels hinder their effective fine-tuning. To address these challenges, we propose MambaX-Net, a novel semi-supervised, dual-scan 3D segmentation architecture that computes the segmentation for time point t by leveraging the MRI and the corresponding segmentation mask from the previous time point. We introduce two new components: (i) a Mamba-enhanced Cross-Attention Module, which integrates the Mamba block into cross attention to efficiently capture temporal evolution and long-range spatial dependencies, and (ii) a Shape Extractor Module that encodes the previous segmentation mask into a latent anatomical representation for refined zone delination. Moreover, we introduce a semi-supervised self-training strategy that leverages pseudo-labels generated from a pre-trained nnU-Net, enabling effective learning without expert annotations. MambaX-Net was evaluated on a longitudinal AS dataset, and results showed that it significantly outperforms state-of-the-art U-Net and Transformer-based models, achieving superior prostate zone segmentation even when trained on limited and noisy data.

Existing benchmarks for large language models (LLMs) are largely restricted to high- or mid-resource languages, and often evaluate performance on higher-order tasks in reasoning and generation. However, plenty of evidence points to the fact that LLMs lack basic linguistic competence in the vast majority of the world's 3800+ written languages. We introduce ChiKhaPo, consisting of 8 subtasks of varying difficulty designed to evaluate the lexical comprehension and generation abilities of generative models. ChiKhaPo draws on existing lexicons, monolingual data, and bitext, and provides coverage for 2700+ languages for 2 subtasks, surpassing any existing benchmark in terms of language coverage. We further show that 6 SOTA models struggle on our benchmark, and discuss the factors contributing to performance scores, including language family, language resourcedness, task, and comprehension versus generation directions. With ChiKhaPo, we hope to enable and encourage the massively multilingual benchmarking of LLMs.

Designing effective agentic systems requires the seamless composition and integration of agents, tools, and models within dynamic and uncertain environments. Most existing methods rely on static, semantic retrieval approaches for tool or agent discovery. However, effective reuse and composition of existing components remain challenging due to incomplete capability descriptions and the limitations of retrieval methods. Component selection suffers because the decisions are not based on capability, cost, and real-time utility. To address these challenges, we introduce a structured, automated framework for agentic system composition that is inspired by the knapsack problem. Our framework enables a composer agent to systematically identify, select, and assemble an optimal set of agentic components by jointly considering performance, budget constraints, and compatibility. By dynamically testing candidate components and modeling their utility in real-time, our approach streamlines the assembly of agentic systems and facilitates scalable reuse of resources. Empirical evaluation with Claude 3.5 Sonnet across five benchmarking datasets shows that our online-knapsack-based composer consistently lies on the Pareto frontier, achieving higher success rates at significantly lower component costs compared to our baselines. In the single-agent setup, the online knapsack composer shows a success rate improvement of up to 31.6% in comparison to the retrieval baselines. In multi-agent systems, the online knapsack composer increases success rate from 37% to 87% when agents are selected from an agent inventory of 100+ agents. The substantial performance gap confirms the robust adaptability of our method across diverse domains and budget constraints.

Reinforcement learning has powered many of the recent breakthroughs in large language models, especially for tasks where rewards can be computed automatically, such as code generation. However, these methods deteriorate in open-ended domains like medical consultation, where feedback is inherently ambiguous, highly context-dependent, and cannot be reduced to a reliable scalar signal. In such settings, RL must either rely on supervision-intensive reward models that often fail to generalize, or it falls into pathological behaviors such as reward hacking - an especially troubling risk for high-stakes medical dialogue. To address these limitations, we introduce ORBIT, an open-ended rubric-based incremental training framework for high-stakes medical dialogue. ORBIT integrates synthetic dialogue generation with dynamically constructed rubrics that serve as adaptive guides for incremental RL. Instead of relying on external medical knowledge bases or handcrafted rule sets, ORBIT uses rubric-driven feedback to steer the learning process. Its judge component can be instantiated with general-purpose instruction-following LLMs, removing the need for any task-specific fine-tuning. Applied to the Qwen3-4B-Instruct model, ORBIT raises the HealthBench-Hard score from 7.0 to 27.5 using only 2k training samples, achieving SOTA performance for models at this scale. With larger rubric datasets, ORBIT-trained models further compete with the strongest open-source baselines on HealthBench-Hard. Our analysis shows that rubric-guided RL consistently improves consultation quality across diverse medical scenarios. We also apply such rubric generation and training pipeline to InfoBench, where ORBIT enhances instruction-following performance, highlighting the generality of rubric-based feedback.

Early approaches used au-toregressive sequence modeling to map speech or text to motion sequences [19, 9], while diffusion-based generators now dominate for their ability to produce detailed, temporally consistent, and natural gestures [12, 10]. Other works explore discrete motion representations, enabling more controllable synthesis [8]. These models accept either speech or text as input and typically rely on speaker embeddings for multi-speaker modeling, which limits their generalization ability to speakers unseen during training. In contrast, Gelina generates both speech and gestures directly from text, and can also clone voice and gestural style through sequence continuation using a speech-gesture prompt, without relying on speaker embeddings. T ext-to-speech approaches: Lately, TTS has shifted toward data-driven methods, with notable advances in discrete code modeling [4, 5, 6].

Coarse-graining (CG) enables molecular dynamics (MD) simulations of larger systems and longer timescales that are otherwise infeasible with atomistic models. Machine learning potentials (MLPs), with their capacity to capture many-body interactions, can provide accurate approximations of the potential of mean force (PMF) in CG models. Current CG MLPs are typically trained in a bottom-up manner via force matching, which in practice relies on configurations sampled from the unbiased equilibrium Boltzmann distribution to ensure thermodynamic consistency. This convention poses two key limitations: first, sufficiently long atomistic trajectories are needed to reach convergence; and second, even once equilibrated, transition regions remain poorly sampled. To address these issues, we employ enhanced sampling to bias along CG degrees of freedom for data generation, and then recompute the forces with respect to the unbiased potential. This strategy simultaneously shortens the simulation time required to produce equilibrated data and enriches sampling in transition regions, while preserving the correct PMF. We demonstrate its effectiveness on the Müller-Brown potential and capped alanine, achieving notable improvements. Our findings support the use of enhanced sampling for force matching as a promising direction to improve the accuracy and reliability of CG MLPs.

Large language models have advanced natural language understanding and generation, but their use as autonomous agents introduces architectural challenges for multi-step tasks. Existing frameworks often mix cognition, memory, and control in a single prompt, reducing coherence and predictability. The Structured Cognitive Loop (SCL) is proposed as an alternative architecture that separates these functions. In SCL, the language model handles cognition, memory is stored externally, and execution is guided by a lightweight controller within a goal-directed loop. This design allows intermediate results to be recorded and verified before actions are taken, improving traceability and evaluation. SCL is evaluated against prompt-based baselines such as ReAct and LangChain agents across three tasks: travel planning, conditional email drafting, and constraint-guided image generation. Under matched settings, SCL achieves an average task success rate of 86.3 percent, compared with 70.5 to 76.8 percent for baselines. It also shows higher goal fidelity, fewer redundant calls, and reduced unsupported assertions. These results indicate that separating cognition, memory, and control can enhance reliability and interpretability without relying on larger models or heavier prompts. The findings should be regarded as preliminary evidence, with broader tests across model families and task domains planned for future work.

Computer-using agents (CUAs) must plan task workflows across diverse and evolving applications, yet progress is limited by the lack of large-scale, high-quality training data. Existing datasets are narrow, static, and costly to annotate, while synthetic data often yields oversimplified or misaligned behaviors. We present Watch & Learn (W&L), a framework that converts readily available Internet videos of human computer use into executable UI trajectories at scale. Instead of directly generating actions or relying on handcrafted heuristics, we cast trajectory annotation as an inverse dynamics problem that predicts user actions from consecutive screen states, which simplifies learning and generalizes across domains. Through a task-aware retrieval and labeling pipeline, W&L yields over 53K high-quality trajectories that enhance CUAs both as in-context exemplars and as supervised training data. On OSWorld, it consistently improves general-purpose and specialized CUAs, while on WindowsAgentArena it achieves state-of-the-art performance among 7B-scale models under the 15-step limit. These results show that web-scale human demonstration videos can serve as a practical and scalable foundation for advancing real-world CUAs.

Interchanges are crucial nodes for vehicle transfers between highways, yet the lack of real-time ramp detectors creates blind spots in traffic prediction. To address this, we propose a Spatio-Temporal Decoupled Autoencoder (STDAE), a two-stage framework that leverages cross-modal reconstruction pretraining. In the first stage, STDAE reconstructs historical ramp flows from mainline data, forcing the model to capture intrinsic spatio-temporal relations. Its decoupled architecture with parallel spatial and temporal autoencoders efficiently extracts heterogeneous features. In the prediction stage, the learned representations are integrated with models such as GWNet to enhance accuracy. Experiments on three real-world interchange datasets show that STDAE-GWNET consistently outperforms thirteen state-of-the-art baselines and achieves performance comparable to models using historical ramp data. This demonstrates its effectiveness in overcoming detector scarcity and its plug-and-play potential for diverse forecasting pipelines.

Today's LLM ecosystem comprises a wide spectrum of models that differ in size, capability, and cost. No single model is optimal for all scenarios; hence, LLM routers have become essential for selecting the most appropriate model under varying circumstances. However, the rapid emergence of various routers makes choosing the right one increasingly challenging. To address this problem, we need a comprehensive router comparison and a standardized leaderboard, similar to those available for models. In this work, we introduce RouterArena, the first open platform enabling comprehensive comparison of LLM routers. RouterArena has (1) a principally constructed dataset with broad knowledge domain coverage, (2) distinguishable difficulty levels for each domain, (3) an extensive list of evaluation metrics, and (4) an automated framework for leaderboard updates. Leveraging our framework, we have produced the initial leaderboard with detailed metrics comparison as shown in Figure 1. Our framework for evaluating new routers is on https://github.com/RouteWorks/RouterArena. Our leaderboard is on https://routeworks.github.io/.