deep-neural-networks

Documents about deep-neural-networks

• QuantSightBench: Evaluating LLM Quantitative Forecasting with Prediction Intervals#31297arXiv CS
• Reversible Residual Normalization Alleviates Spatio-Temporal Distribution Shift#31298arXiv CS
• Supplement Generation Training for Enhancing Agentic Task Performance#124052arXiv CS
• Tokenised Flow Matching for Hierarchical Simulation Based Inference#124053arXiv CS
• COMPASS: COntinual Multilingual PEFT with Adaptive Semantic Sampling#124054arXiv CS
• Too Sharp, Too Sure: When Calibration Follows Curvature#124064arXiv CS
• On Bayesian Softmax-Gated Mixture-of-Experts Models#124070arXiv CS
• CHASM: Unveiling Covert Advertisements on Chinese Social Media#124072arXiv CS
• Fine-Tuning Regimes Define Distinct Continual Learning Problems#126495arXiv CS
• PrismaDV: Automated Task-Aware Data Unit Test Generation#126511arXiv CS
• Transferable SCF-Acceleration through Solver-Aligned Initialization Learning#126524arXiv CS
• A Kernel Nonconformity Score for Multivariate Conformal Prediction#126532arXiv CS
• CoFEE: Reasoning Control for LLM-Based Feature Discovery#126533arXiv CS
• Adaptive Head Budgeting for Efficient Multi-Head Attention#134544arXiv CS
• Useful nonrobust features are ubiquitous in biomedical images#134546arXiv CS
• Enhancing molecular dynamics with equivariant machine-learned densities#138927arXiv CS
• MIMIC: A Generative Multimodal Foundation Model for Biomolecules#138938arXiv CS
• Modeling Behavioral Intensity and Transitions for Generative Recommendation#138941arXiv CS
• Scaling Properties of Continuous Diffusion Spoken Language Models#138944arXiv CS
• An Aircraft Upset Recovery System with Reinforcement Learning#138953arXiv CS
• Investigation into In-Context Learning Capabilities of Transformers#141474arXiv CS
• Towards interpretable AI with quantum annealing feature selection#141486arXiv CS
• Safe-Support Q-Learning: Learning without Unsafe Exploration#141500arXiv CS
• Online learning with Erdős-Rényi side-observation graphs#141509arXiv CS
• PhyCo: Learning Controllable Physical Priors for Generative Motion#146912arXiv CS
• Global Optimality for Constrained Exploration via Penalty Regularization#146917arXiv CS
• Shuffling-Aware Optimization for Private Vector Mean Estimation#146931arXiv CS
• Probabilistic Circuits for Irregular Multivariate Time Series Forecasting#146955arXiv CS
• Generating Statistical Charts with Validation-Driven LLM Workflows#151715arXiv CS
• Position: agentic AI orchestration should be Bayes-consistent#151725arXiv CS
• Temporal Data Requirement for Predicting Unplanned Hospital Readmissions#151727arXiv CS
• Knowing when to trust machine-learned interatomic potentials#151745arXiv CS
• Jailbreaking Vision-Language Models Through the Visual Modality#151753arXiv CS
• Gradient Regularized Newton Boosting Trees with Global Convergence#151754arXiv CS
• Exact ReLU realization of tensor-product refinement iterates#155282arXiv CS
• Agentic-imodels: Evolving agentic interpretability tools via autoresearch#155297arXiv CS
• GEM-FI: Gated Evidential Mixtures with Fisher Modulation#155304arXiv CS
• Rethinking the Rank Threshold for LoRA Fine-Tuning#155307arXiv CS
• Verifier-Backed Hard Problem Generation for Mathematical Reasoning#168299arXiv CS
• Crafting Reversible SFT Behaviors in Large Language Models#168309arXiv CS
• Cross-Modal Navigation with Multi-Agent Reinforcement Learning#168322arXiv CS
• ReActor: Reinforcement Learning for Physics-Aware Motion Retargeting#168323arXiv CS
• Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning#175241arXiv CS
• Equivariant Reinforcement Learning for Clifford Quantum Circuit Synthesis#175243arXiv CS
• DataMaster: Towards Autonomous Data Engineering for Machine Learning#175245arXiv CS
• Compute Where it Counts: Self Optimizing Language Models#175253arXiv CS
• Policy Gradient Methods for Non-Markovian Reinforcement Learning#175266arXiv CS
• PhyGround: Benchmarking Physical Reasoning in Generative World Models#175269arXiv CS
• Elucidating Representation Degradation Problem in Diffusion Model Training#175275arXiv CS
• Learning Minimally Rigid Graphs with High Realization Counts#178875arXiv CS