Automated feedback generation plays a crucial role in enhancing personalized learning experiences in computer science education. Among different types of feedback, next-step hint feedback is particularly important, as it provides students with actionable steps to progress towards solving programming tasks. This study investigates how students interact with an AI-driven next-step hint system in an in-IDE learning environment. We gathered and analyzed a dataset from 34 students solving Kotlin tasks, containing detailed hint interaction logs. We applied process mining techniques and identified 16 common interaction scenarios. Semi-structured interviews with 6 students revealed strategies for managing unhelpful hints, such as adapting partial hints or modifying code to generate variations of the same hint. These findings, combined with our publicly available dataset, offer valuable opportunities for future research and provide key insights into student behavior, helping improve hint design for enhanced learning support.

Recent reasoning-first models (e.g., OpenAI o1, DeepSeek R1) have spurred a resurgence of interest in RLVR. Nevertheless, advances are dominated by mathematics (e.g., AIME), with competitive-programming code generation underexplored and data curation receiving less attention than RL algorithm design. We investigate how to construct RLVR datasets (i.e., RL prompts) and present practical training techniques that yield strong performance on competitive-programming code generation. Our pipeline begins with supervised fine-tuning (SFT) distilled from strong open-source models, augmented with general-purpose and reasoning-intensive data. RL then follows a two-stage process with executable, testcase-driven rewards: first, training on a large, uniformly distributed set of competitive-programming problems using Group Relative Policy Optimization (GRPO) with 8 rollouts per prompt and a relatively short response-generation window (e.g., 32k during SFT and 24k in this stage) to expand entropy and mitigate repetition and truncation; second, we perform \textbf{Pre-GRPO}: updating on a small, high-quality set of challenging problems with a large rollout budget (64 rollouts per prompt) under a hard-focus curriculum that continuously retains the most difficult instances throughout training. We implement our method on Qwen2.5-32B and evaluate on LeetCode and Codeforces weekly contests to avoid data leakage. The resulting model achieves state-of-the-art performance among models of similar scale and is comparable to leading systems such as DeepSeek v3.1 and Doubao-1.5-Thinking. We also examine scaling trends and observe strong RL scaling on an internal large-scale MoE model. Our study distills concise best practices for data curation, entropy expansion, and curriculum design in RLVR for competitive-programming code generation.

Adapting Large Language Models (LLMs) to a continuous stream of tasks is a critical yet challenging endeavor. While Parameter-Efficient Fine-Tuning (PEFT) methods have become a standard for this, they face a fundamental dilemma in continual learning. Reusing a single set of PEFT parameters for new tasks often leads to catastrophic forgetting of prior knowledge. Conversely, allocating distinct parameters for each task prevents forgetting but results in a linear growth of the model's size and fails to facilitate knowledge transfer between related tasks. To overcome these limitations, we propose a novel adaptive PEFT method referred to as \textit{Mixtures of SubExperts (MoSEs)}, a novel continual learning framework designed for minimal forgetting and efficient scalability. MoSEs integrate a sparse Mixture of SubExperts into the transformer layers, governed by a task-specific routing mechanism. This architecture allows the model to isolate and protect knowledge within dedicated SubExperts, thereby minimizing parameter interference and catastrophic forgetting. Crucially, the router can adaptively select and combine previously learned sparse parameters for new tasks, enabling effective knowledge transfer while ensuring that the model's capacity grows sublinearly. We evaluate MoSEs on the comprehensive TRACE benchmark datasets. Our experiments demonstrate that MoSEs significantly outperform conventional continual learning approaches in both knowledge retention and scalability to new tasks, achieving state-of-the-art performance with substantial memory and computational savings.

We study architectural and optimization techniques for sample-efficient language modeling under the constraints of the BabyLM 2025 shared task. Our model, BLaLM, replaces self-attention with a linear-time mLSTM token mixer and explores lightweight enhancements, including short convolutions, sliding window attention with dynamic modulation, and Hedgehog feature maps. To support training in low-resource settings, we curate a high-quality corpus emphasizing readability and pedagogical structure. Experiments across both STRICT and STRICT-SMALL tracks show that (1) linear attention combined with sliding window attention consistently improves zero-shot performance, and (2) the Muon optimizer stabilizes convergence and reduces perplexity over AdamW. These results highlight effective strategies for efficient language modeling without relying on scale.

While reinforcement learning (RL) can empower autonomous agents by enabling self-improvement through interaction, its practical adoption remains challenging due to costly rollouts, limited task diversity, unreliable reward signals, and infrastructure complexity, all of which obstruct the collection of scalable experience data. To address these challenges, we introduce DreamGym, the first unified framework designed to synthesize diverse experiences with scalability in mind to enable effective online RL training for autonomous agents. Rather than relying on expensive real-environment rollouts, DreamGym distills environment dynamics into a reasoning-based experience model that derives consistent state transitions and feedback signals through step-by-step reasoning, enabling scalable agent rollout collection for RL. To improve the stability and quality of transitions, DreamGym leverages an experience replay buffer initialized with offline real-world data and continuously enriched with fresh interactions to actively support agent training. To improve knowledge acquisition, DreamGym adaptively generates new tasks that challenge the current agent policy, enabling more effective online curriculum learning. Experiments across diverse environments and agent backbones demonstrate that DreamGym substantially improves RL training, both in fully synthetic settings and in sim-to-real transfer scenarios. On non-RL-ready tasks like WebArena, DreamGym outperforms all baselines by over 30%. And in RL-ready but costly settings, it matches GRPO and PPO performance using only synthetic interactions. When transferring a policy trained purely on synthetic experiences to real-environment RL, DreamGym yields significant additional performance gains while requiring far fewer real-world interactions, providing a scalable warm-start strategy for general-purpose RL.

Autonomous agents are increasingly expected to operate in complex, dynamic, and uncertain environments, performing tasks such as manipulation, navigation, and decision-making. Achieving these capabilities requires agents to understand the underlying mechanisms and dynamics of the world, moving beyond reactive control or simple replication of observed states. This motivates the development of world models as internal representations that encode environmental states, capture dynamics, and support prediction, planning, and reasoning. Despite growing interest, the definition, scope, architectures, and essential capabilities of world models remain ambiguous. In this survey, we go beyond prescribing a fixed definition and limiting our scope to methods explicitly labeled as world models. Instead, we examine approaches that exhibit the core capabilities of world models through a review of methods in robotic manipulation. We analyze their roles across perception, prediction, and control, identify key challenges and solutions, and distill the core components, capabilities, and functions that a fully realized world model should possess. Building on this analysis, we aim to motivate further development toward generalizable and practical world models for robotics.

Abstract--Integrating generative models with action chunking has shown significant promise in imitation learning for robotic manipulation. However, the existing diffusion-based paradigm often struggles to capture strong temporal dependencies across multiple steps, particularly when incorporating proprioceptive input. This limitation can lead to task failures, where the policy overfits to proprioceptive cues at the expense of capturing the visually derived features of the task. T o overcome this challenge, we propose the Deep Koopman-boosted Dual-branch Diffusion Policy (D3P) algorithm. D3P introduces a dual-branch architecture to decouple the roles of different sensory modality combinations. The visual branch encodes the visual observations to indicate task progression, while the fused branch integrates both visual and proprioceptive inputs for precise manipulation. Within this architecture, when the robot fails to accomplish intermediate goals, such as grasping a drawer handle, the policy can dynamically switch to execute action chunks generated by the visual branch, allowing recovery to previously observed states and facilitating retrial of the task. T o further enhance visual representation learning, we incorporate a Deep Koopman Operator module that captures structured temporal dynamics from visual inputs. During inference, we use the test-time loss of the generative model as a confidence signal to guide the aggregation of the temporally overlapping predicted action chunks, thereby enhancing the reliability of policy execution. In simulation experiments across six RLBench tabletop tasks, D3P outperforms the state-of-the-art diffusion policy by an average of 14.6%. On three real-world robotic manipulation tasks, it achieves a 15.0% improvement.

Recent LLM benchmarks have tested models on a range of phenomena, but are still focused primarily on natural language understanding for extraction of explicit information, such as QA or summarization, with responses often targeting information from individual sentences. We are still lacking more challenging, and importantly also multilingual, benchmarks focusing on implicit information and pragmatic inferences across larger documents in the context of discourse tracking: integrating and aggregating information across sentences, paragraphs and multiple speaker utterances. To this end, we present DiscoTrack, an LLM benchmark targeting a range of tasks across 12 languages and four levels of discourse understanding: salience recognition, entity tracking, discourse relations and bridging inference. Our evaluation shows that these tasks remain challenging, even for state-of-the-art models.

We revisit the problem of private online learning, in which a learner receives a sequence of $T$ data points and has to respond at each time-step a hypothesis. It is required that the entire stream of output hypotheses should satisfy differential privacy. Prior work of Golowich and Livni [2021] established that every concept class $\mathcal{H}$ with finite Littlestone dimension $d$ is privately online learnable in the realizable setting. In particular, they proposed an algorithm that achieves an $O_{d}(\log T)$ mistake bound against an oblivious adversary. However, their approach yields a suboptimal $\tilde{O}_{d}(\sqrt{T})$ bound against an adaptive adversary. In this work, we present a new algorithm with a mistake bound of $O_{d}(\log T)$ against an adaptive adversary, closing this gap. We further investigate the problem in the agnostic setting, which is more general than the realizable setting as it does not impose any assumptions on the data. We give an algorithm that obtains a sublinear regret of $\tilde{O}_d(\sqrt{T})$ for generic Littlestone classes, demonstrating that they are also privately online learnable in the agnostic setting.

We explore the effectiveness and reliability of an artificial intelligence (AI)-based grading system for a handwritten general chemistry exam, comparing AI-assigned scores to human grading across various types of questions. Exam pages and grading rubrics were uploaded as images to account for chemical reaction equations, short and long open-ended answers, numerical and symbolic answer derivations, drawing, and sketching in pencil-and-paper format. Using linear regression analyses and psychometric evaluations, the investigation reveals high agreement between AI and human graders for textual and chemical reaction questions, while highlighting lower reliability for numerical and graphical tasks. The findings emphasize the necessity for human oversight to ensure grading accuracy, based on selective filtering. The results indicate promising applications for AI in routine assessment tasks, though careful consideration must be given to student perceptions of fairness and trust in integrating AI-based grading into educational practice.