Reinforcement learning (RL) has emerged as an effective paradigm for enhancing model reasoning. However, existing RL methods like GRPO often rely on unstructured self-sampling to fit scalar rewards, often producing inefficient rollouts that fail to capture transferable problem-solving strategies. To address these limitations, we propose **TemplateRL**, a structured template-guided RL framework that augments policy optimization with explicit template guidance. Our approach first constructs a problem-solving template library via MCTS on a small seed set, then seamlessly integrates this high-level structured guidance into RL training. By guiding rollout generation to align with proven template structures, TemplateRL significantly improves high-quality trajectory hit rates while reducing ineffective exploration. This structure-guided design steers the policy toward validated strategic patterns, stabilizing training dynamics, and enhancing RL sampling efficiency. Notably, the explicit template library is interpretable, editable, and supports online updates-enabling continuous updates during both training and inference. Extensive experiments demonstrate that TemplateRL outperforms GRPO by 99% on AIME and 41% on AMC, with superior stability on weak models and remarkable cross-domain generalization, highlighting its potential for broader tasks.

Large Language Models (LLMs) have recently achieved remarkable performance in long-context understanding. However, current long-context LLM benchmarks are limited by rigid context length, labor-intensive annotation, and the pressing challenge of label leakage issues during LLM training. Therefore, we propose \textsc{AcademicEval}, a live benchmark for evaluating LLMs over long-context generation tasks. \textsc{AcademicEval} adopts papers on arXiv to introduce several academic writing tasks with long-context inputs, \textit{i.e.}, \textsc{Title}, \textsc{Abstract}, \textsc{Introduction}, and \textsc{Related Work}, which cover a wide range of abstraction levels and require no manual labeling. Moreover, \textsc{AcademicEval} integrates high-quality and expert-curated few-shot demonstrations from a collected co-author graph to enable flexible context length. Especially, \textsc{AcademicEval} features an efficient live evaluation, ensuring no label leakage. We conduct a holistic evaluation on \textsc{AcademicEval}, and the results illustrate that LLMs perform poorly on tasks with hierarchical abstraction levels and tend to struggle with long few-shot demonstrations, highlighting the challenge of our benchmark. Through experimental analysis, we also reveal some insights for enhancing LLMs' long-context modeling capabilities. Code is available at https://github.com/ulab-uiuc/AcademicEval

Modeling dynamical systems and unraveling their underlying causal relationships is central to many domains in the natural sciences. Various physical systems, such as those arising in cell biology, are inherently high-dimensional and stochastic in nature, and admit only partial, noisy state measurements. This poses a significant challenge for addressing the problems of modeling the underlying dynamics and inferring the network structure of these systems. Existing methods are typically tailored either for structure learning or modeling dynamics at the population level, but are limited in their ability to address both problems together. In this work, we address both problems simultaneously: we present StructureFlow, a novel and principled simulation-free approach for jointly learning the structure and stochastic population dynamics of physical systems. We showcase the utility of StructureFlow for the tasks of structure learning from interventions and dynamical (trajectory) inference of conditional population dynamics. We empirically evaluate our approach on high-dimensional synthetic systems, a set of biologically plausible simulated systems, and an experimental single-cell dataset. We show that StructureFlow can learn the structure of underlying systems while simultaneously modeling their conditional population dynamics -- a key step toward the mechanistic understanding of systems behavior.

The remarkable success of machine learning (ML) in predictive tasks has led scientists to incorporate ML predictions as a core component of the scientific discovery pipeline. This was exemplified by the landmark achievement of AlphaFold (Jumper et al. (2021)). In this paper, we study how ML predictions can be safely used in statistical analysis of data towards scientific discovery. In particular, we follow the framework introduced by Angelopoulos et al. (2023). In this framework, we assume access to a small set of $n$ gold-standard labeled samples, a much larger set of $N$ unlabeled samples, and a ML model that can be used to impute the labels of the unlabeled data points. We introduce two new learning-augmented estimators: (1) Prediction-Augmented Residual Tree (PART), and (2) Prediction-Augmented Quadrature (PAQ). Both estimators have significant advantages over existing estimators like PPI and PPI++ introduced by Angelopoulos et al. (2023) and Angelopoulos et al. (2024), respectively. PART is a decision-tree based estimator built using a greedy criterion. We first characterize PART's asymptotic distribution and demonstrate how to construct valid confidence intervals. Then we show that PART outperforms existing methods in real-world datasets from ecology, astronomy, and census reports, among other domains. This leads to estimators with higher confidence, which is the result of using both the gold-standard samples and the machine learning predictions. Finally, we provide a formal proof of the advantage of PART by exploring PAQ, an estimation that arises when considering the limit of PART when the depth its tree grows to infinity. Under appropriate assumptions in the input data we show that the variance of PAQ shrinks at rate of $O(N^{-1} + n^{-4})$, improving significantly on the $O(N^{-1}+n^{-1})$ rate of existing methods.

We introduce a novel regularization scheme for autoencoders based on matricial free energy. Our approach defines a differentiable loss function in terms of the singular values of the code matrix (code dimension x batch size). From the standpoint of free probability an d random matrix theory, this loss achieves its minimum when the singular value distribution of the code matrix coincides with that of an appropriately sculpted random metric with i.i.d. Gaussian entries. Empirical simulations demonstrate that minimizing the negative matricial free energy through standard stochastic gradient-based training yields Gaussian-like codes that generalize across training and test sets. Building on this foundation, we propose a matricidal free energy maximizing autoencoder that reliably produces Gaussian codes and show its application to underdetermined inverse problems.

Real-world multivariate time series anomalies are rare and often unlabeled. Additionally, prevailing methods rely on increasingly complex architectures tuned to benchmarks, detecting only fragments of anomalous segments and overstating performance. In this paper, we introduce OracleAD, a simple and interpretable unsupervised framework for multivariate time series anomaly detection. OracleAD encodes each variable's past sequence into a single causal embedding to jointly predict the present time point and reconstruct the input window, effectively modeling temporal dynamics. These embeddings then undergo a self-attention mechanism to project them into a shared latent space and capture spatial relationships. These relationships are not static, since they are modeled by a property that emerges from each variable's temporal dynamics. The projected embeddings are aligned to a Stable Latent Structure (SLS) representing normal-state relationships. Anomalies are identified using a dual scoring mechanism based on prediction error and deviation from the SLS, enabling fine-grained anomaly diagnosis at each time point and across individual variables. Since any noticeable SLS deviation originates from embeddings that violate the learned temporal causality of normal data, OracleAD directly pinpoints the root-cause variables at the embedding level. OracleAD achieves state-of-the-art results across multiple real-world datasets and evaluation protocols, while remaining interpretable through SLS.

The intricate dynamics of stock markets have led to extensive research on models that are able to effectively explain their inherent complexities. This study leverages the econometrics literature to explore the dynamic factor model as an interpretable model with sufficient predictive capabilities for capturing essential market phenomena. Although the model has been extensively applied for predictive purposes, this study focuses on analyzing the extracted loadings and common factors as an alternative framework for understanding stock price dynamics. The results reveal novel insights into traditional market theories when applied to the Philippine Stock Exchange using the Kalman method and maximum likelihood estimation, with subsequent validation against the capital asset pricing model. Notably, a one-factor model extracts a common factor representing systematic or market dynamics similar to the composite index, whereas a two-factor model extracts common factors representing market trends and volatility. Furthermore, an application of the model for nowcasting the growth rates of the Philippine gross domestic product highlights the potential of the extracted common factors as viable real-time market indicators, yielding over a 34% decrease in the out-of-sample prediction error. Overall, the results underscore the value of dynamic factor analysis in gaining a deeper understanding of market price movement dynamics.

Existing expert merging strategies for Sparse Mixture of Experts (SMoE) typically rely on input-dependent or input-independent averaging of expert parameters, but often lack a principled weighting mechanism. In this work, we reinterpret expert merging through the lens of game theory, revealing cooperative and competitive dynamics among experts. Based on this perspective, we introduce Nash Merging of Experts (NAMEx), a novel framework that incorporates Nash Bargaining into the merging process, enabling more balanced and efficient collaboration among experts. Additionally, we incorporate complex momentum into NAMEx to accelerate expert propagation with theoretical guarantees for convergence. Extensive experiments across language modelling, text classification, image classification, and zero-shot robustness under data corruption show that NAMEx consistently outperforms competing methods while integrating seamlessly with popular MoE architectures. Finally, we demonstrate NAMEx's scalability by applying it to large-scale systems, including Qwen1.5-MoE (14B) and DeepSeek-MoE (16B), where it proves effective in both zero-shot and fine-tuning settings.

Modeling irregularly sampled multivariate time series is a persistent challenge in domains like healthcare and sensor networks. While recent works have explored a variety of complex learning architectures to solve the prediction problems for irregularly sampled time series, it remains unclear what are the true benefits of some of these architectures, and whether clever modifications of simpler and more efficient RNN-based algorithms are still competitive, i.e. they are on par with or even superior to these methods. In this work, we propose and study GRUwE: Gated Recurrent Unit with Exponential basis functions, that builds upon RNN-based architectures for observations made at irregular times. GRUwE supports both regression-based and event-based predictions in continuous time. GRUwE works by maintaining a Markov state representation of the time series that updates with the arrival of irregular observations. The Markov state update relies on two reset mechanisms: (i) observation-triggered reset, and (ii) time-triggered reset of the GRU state using learn-able exponential decays, to support the predictions in continuous time. Our empirical evaluations across several real-world benchmarks on next-observation and next-event prediction tasks demonstrate that GRUwE can indeed achieve competitive to superior performance compared to the recent state-of-the-art (SOT A) methods. Thanks to its simplicity, GRUwE offers compelling advantages: it is easy to implement, requires minimal hyper-parameter tuning efforts, and significantly reduces the computational overhead in the online deployment.

Existing benchmarks for evaluating mathematical reasoning in large language models (LLMs) rely primarily on competition problems, formal proofs, or artificially challenging questions -- failing to capture the nature of mathematics encountered in actual research environments. We introduce RealMath, a novel benchmark derived directly from research papers and mathematical forums that assesses LLMs' abilities on authentic mathematical tasks. Our approach addresses three critical challenges: sourcing diverse research-level content, enabling reliable automated evaluation through verifiable statements, and designing a continually refreshable dataset to mitigate contamination risks. Experimental results across multiple LLMs reveal surprising capabilities in handling research mathematics compared to competition problems, suggesting current models may already serve as valuable assistants for working mathematicians despite limitations on highly challenging problems. The code and dataset for RealMath are publicly available.