arXiv:2604.15460v1 Announce Type: new Abstract: The rapid evolution of Large Language Models (LLMs) has made them powerful tools for enhancing student writing. This study explores the extent and limitations of LLMs in assisting secondary-level English as a Foreign Language (EFL) students with their writing tasks. While existing studies focus on output quality, our research examines the developmental shift in LLMs and their impact on EFL students, assessing whether smarter models act as true scaffolds or mere compensatory crutches. […]

Ver mais

Continual learning, enabling models to acquire new skills and knowledge without degrading existing capabilities, remains a fundamental challenge for foundation models. While on-policy reinforcement learning can reduce forgetting, it requires explicit reward functions that are often unavailable. Learning from expert demonstrations, the primary alternative, is dominated by supervised fine-tuning (SFT), which is inherently off-policy. We introduce Self-Distillation Fine-Tuning (SDFT), a simple method that enables on-policy learning directly from demonstrations. SDFT leverages in-context learning by using a demonstration-conditioned model […]

Ver mais

arXiv:2603.09310v1 Announce Type: cross Abstract: We study training algorithms with data following a Gaussian mixture model. For a specific family of such algorithms, we present a non-asymptotic result, connecting the evolution of the model to a surrogate dynamical system, which can be easier to analyze. The proof of our result is based on the celebrated Gordon comparison theorem. Using our theorem, we rigorously prove the validity of the dynamic mean-field (DMF) expressions in the asymptotic scenarios. Moreover, we […]

Ver mais

arXiv:2603.16964v1 Announce Type: new Abstract: Approval of ADS depends on evaluating its behavior within representative real-world traffic scenarios. A common way to obtain such scenarios is to extract them from real-world data recordings. These can then be grouped and serve as basis on which the ADS is subsequently tested. This poses two central challenges: how scenarios are extracted and how they are grouped. Existing extraction methods rely on heterogeneous definitions, hindering scenario comparability. For the grouping of scenarios, […]

Ver mais

Deepfake detection from images and videos has evolved from artifact-specific convolutional baselines toward more generalizable, cross-dataset, and foundation-model-based approaches. The current work focuses on the efficiency and informativeness of frame selection itself, while keeping the downstream detectors fixed. The study compares twelve frame-selection heuristics ranging from simple baselines to landmark-aware strategies. Four pre-trained detectors were included in the present quantitative comparison: Self-Blended Images (SBI), Frequency-Enhanced Self-Lendered Images (FSBI), Generative Convolutional Vision Transformer (GenConViT), and GenD. The results show […]

Ver mais

arXiv:2602.15074v1 Announce Type: new Abstract: We introduce a structure-aware approach for symbolic piano accompaniment that decouples high-level planning from note-level realization. A lightweight transformer predicts an interpretable, per-measure style plan conditioned on section/phrase structure and functional harmony, and a retriever then selects and reharmonizes human-performed piano patterns from a corpus. We formulate retrieval as pattern matching under an explicit energy with terms for harmonic feasibility, structural-role compatibility, voice-leading continuity, style preferences, and repetition control. Given a structured lead […]

Ver mais

Reinforcement Learning from Verifiable Rewards (RLVR) significantly enhances large language models (LLMs) reasoning but severely suffers from calibration degeneration, where models become excessively over-confident in incorrect answers. Previous studies devote to directly incorporating calibration objective into existing optimization target. However, our theoretical analysis demonstrates that there exists a fundamental gradient conflict between the optimization for maximizing policy accuracy and minimizing calibration error. Building on this insight, we propose DCPO, a simple yet effective framework that systematically decouples reasoning […]

Ver mais

Cadence Design Systems announced two AI-related collaborations at its CadenceLIVE event this week, expanding its work with Nvidia and introducing new integrations with Google Cloud. The Nvidia partnership focuses on combining AI with physics-based simulation and accelerated computing for robotic systems and system-level design. The companies said the approach targets modelling and deployment in semiconductors and large-scale AI infrastructure, including robotic systems that Nvidia describes as physical AI. Cadence is integrating its multi-physics simulation and system design tools […]

Ver mais

arXiv:2603.00888v1 Announce Type: cross Abstract: Despite exceptional predictive performance of Deep sequence models (DSMs), the main concern of their deployment centers around the lack of uncertainty awareness. In contrast, probabilistic models quantify the uncertainty associated with unobserved variables with rules of probability. Notably, Bayesian methods leverage Bayes’ rule to express our belief of unobserved variables in a principled way. Since exact Bayesian inference is computationally infeasible at scale, approximate inference is required in practice. Two major bottlenecks of […]

Ver mais

Sparsity is a well-studied technique for compressing deep neural networks (DNNs) without compromising performance. In deep reinforcement learning (DRL), neural networks with up to 5% of their original weights can still be trained with minimal performance loss compared to their dense counterparts. However, most existing methods rely on unstructured fine-grained sparsity, which limits hardware acceleration opportunities due to irregular computation patterns. Structured coarse-grained sparsity enables hardware acceleration, yet typically degrades performance and increases pruning complexity. In this work, […]

Ver mais