arXiv:2601.20886v1 Announce Type: new Abstract: IDE-Bench is a comprehensive framework for evaluating AI IDE agents on real-world software engineering tasks through an IDE-native tool interface. We present a Dockerized test harness that goes beyond raw terminal execution, granting models a structured tool ecosystem that represents AI-native IDEs like Cursor and Windsurf. By providing high-level abstractions for codebase search, structured file editing, and tools for testing full-stack applications, IDE-Bench evaluates an agent’s ability to act as a true engineering […]

Ver mais

arXiv:2510.03871v2 Announce Type: replace-cross Abstract: Despite recent progress in optimal hyperparameter transfer under model and dataset scaling, no unifying explanatory principle has been established. For Adam and Scion optimizers, we discover that joint optimal scaling across model and dataset sizes is conditioned on a single invariant: the operator norm of the output layer. Across models with up to 1.3B parameters trained on up to 138B tokens, the optimal learning rate/batch size pair $(eta^{ast}, B^{ast})$ consistently has the same […]

Ver mais

I’m excited to share a major update to sklearn-diagnose – the open-source Python library that acts as an “MRI scanner” for your ML models (https://www.reddit.com/r/MachineLearning/s/EcMRYPVIDX) When I first released sklearn-diagnose, users could generate diagnostic reports to understand why their models were failing. But I kept thinking – what if you could talk to your diagnosis? What if you could ask follow-up questions and drill down into specific issues? Now you can! 🚀 🆕 What’s New: Interactive Diagnostic Chatbot […]

Ver mais

Humanoid robots can suffer significant performance drops under small changes in dynamics, task specifications, or environment setup. We propose HoRD, a two-stage learning framework for robust humanoid control under domain shift. First, we train a high-performance teacher policy via history-conditioned reinforcement learning, where the policy infers latent dynamics context from recent state–action trajectories to adapt online to diverse randomized dynamics. Second, we perform online distillation to transfer the teacher’s robust control capabilities into a transformer-based student policy that […]

Ver mais

Reinforcement learning drives recent advances in LLM reasoning and agentic capabilities, yet current approaches struggle with both exploration and exploitation. Exploration suffers from low success rates on difficult tasks and high costs of repeated rollouts from scratch. Exploitation suffers from coarse credit assignment and training instability: Trajectory-level rewards penalize valid prefixes for later errors, and failure-dominated groups overwhelm the few positive signals, leaving optimization without constructive direction. To this end, we propose R$^3$L, Reflect-then-Retry Reinforcement Learning with Language-Guided […]

Ver mais

Tokenization is a hardcoded compression step which remains in the training pipeline of Large Language Models (LLMs), despite a general trend towards architectures becoming increasingly end-to-end. Prior work has shown promising results at scale in bringing this compression step inside the LLMs’ architecture with heuristics to draw token boundaries, and also attempts to learn these token boundaries with straight-through estimates, which treat the problem of drawing discrete token boundaries as a continuous one. We show that these token […]

Ver mais

arXiv:2601.20071v1 Announce Type: new Abstract: Gradient-regularized value learning methods improve sample efficiency by leveraging learned models of transition dynamics and rewards to estimate return gradients. However, existing approaches, such as MAGE, struggle in stochastic or noisy environments, limiting their applicability. In this work, we address these limitations by extending distributional reinforcement learning on continuous state-action spaces to model not only the distribution over scalar state-action value functions but also over their gradients. We refer to this approach as […]

Ver mais

arXiv:2511.18940v2 Announce Type: replace-cross Abstract: Cross-subject motor-imagery decoding remains a major challenge in EEG-based brain-computer interfaces. To mitigate strong inter-subject variability, recent work has emphasized manifold-based approaches operating on covariance representations. Yet dispersion scaling and orientation alignment remain largely unaddressed in existing methods. In this paper, we address both issues through congruence transforms and introduce three complementary geometry-aware models: (i) Discriminative Congruence Transform (DCT), (ii) Deep Linear DCT (DLDCT), and (iii) Deep DCT-UNet (DDCT-UNet). These models are evaluated […]

Ver mais

arXiv:2601.03281v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly used as high level controllers for autonomous Unmanned Aerial Vehicle (UAV) missions. However, existing evaluations rarely assess whether such agents remain safe, protocol compliant, and effective under realistic next generation networking constraints. This paper introduces $alpha^3$-Bench, a benchmark for evaluating LLM driven UAV autonomy as a multi turn conversational reasoning and control problem operating under dynamic 6G conditions. Each mission is formulated as a language mediated control […]

Ver mais

The rapid proliferation of large language models has created significant opportunities for personalized education, yet existing systems rarely account for user competency as a determinant of interaction quality. This study introduces Persona in The Loop (PITL), a dual-mode adaptive framework that recommends AI personas for blockchain and smart contract education applications. PITL employs 100 AI personas organized across two domains, ten sub-specialties, and five Dreyfus competency levels, recommending personas via either similarity-based mode grounded in Cognitive Load Theory […]

Ver mais