Learning-based method for designing Kazantzis-Kravaris/Luenberger observers for non-autonomous nonlinear systems using hypernetworks for input conditioning.
Reasoning-driven approach for generating synthetic multi-modal training data without manual prompts, addressing scarcity of specialized AI training datasets.
Methods for processing and analyzing petabyte-scale whole-brain 3D microscopy data from light-sheet fluorescence microscopy for neuroscience research.
DIAL proposes decoupling intent and action in Vision-Language-Action models via latent world modeling to improve decision-making and training stability in end-to-end robotic control.
GENIE method for editing Implicit Neural Representations via Gram-eigenmode deformations without retraining.
Hybrid machine learning framework for graduate admission prediction and university-program recommendation using 13,000 GradCafe records.
Method using epistemic uncertainty to identify unreliable explanations in post-hoc XAI methods, reducing explanation generation costs.
Deep learning framework combining segmentation and dual-mode compression for efficient wind turbine inspection imagery transfer.
Uncertainty quantification method for image segmentation using spatially-aware aggregation for out-of-distribution detection.
Proposes adaptive reasoning allocation during code generation for LLMs, addressing limitations of upfront thinking approaches in handling code complexity.
Statistical methods research on Oaxaca-Blinder decomposition showing reference group choice can reverse conclusions about group differences.
Early exiting predictive coding neural networks optimized for edge AI devices with resource constraints and privacy requirements.
GenOL framework for online learning with only concept names (name-only setup) enabling real-time adaptation to data distribution shifts in continual learning scenarios.
Introduces WEATHER-5K dataset and benchmarks physics-informed time-series forecasting models for global weather prediction.
Control-theoretic approach to reinforcement learning with convergence guarantees, new gradient theorem, and gradient ascent algorithm.
Information-theoretic analysis of transformer in-context learning on variable-order Markov chains with finite-sample accuracy bounds.
Diffusion sampler using value functions with invariant symmetries for sampling from unnormalized target densities.
Equivariant neural networks for analyzing scalar and vector fields on spheres using group convolutions in Fourier space.
Critical evaluation of model inversion attack assessment frameworks, identifying flaws in standard evaluation methodology.
Neural Graduated Assignment method for solving Maximum Common Edge Subgraph problem with improved scalability.
Training-free framework for compiling sparse Mixture-of-Experts variants with predicted expert utility metric for deployment optimization.
Theoretical analysis of concentration properties for fractional quasi p-norms in high-dimensional spaces.
Framework for characterizing epistemic errors in uncertainty-aware multitask learners under distribution shift.
Implicit neural representations for efficient exploration of large-scale simulation ensembles with interpretability focus.
Probabilistic inference speedup for Hidden Markov Models by filtering low-probability states in temporal sequences.
Classical polynomial chaos expansion technique for surrogate modeling and uncertainty quantification in physical simulation.
Research on dynamic reward weighting for multi-objective RL alignment in LLMs, addressing non-convex Pareto fronts in preference learning.
Comparative study of neural network classifiers and optimizers for EEG frequency band classification across brain hemispheres.
Theoretical convergence analysis of Muon optimizer for matrix-structured parameters in neural network training.
Deep learning for predicting shock propagation in porous materials with multi-field, spatio-temporal modeling.
Out-of-distribution detection for regression tasks in scientific AI using score-based diffusion models on joint likelihood estimation.
Transformer-based inter-atomic potential model for molecular simulations without explicit equivariance constraints.
Analysis of implicit models with infinite-depth weight-tied networks that match explicit models while reducing memory consumption.
Framework for learning graphon mixtures from graph data using motif moments for clustering graphs from multiple distributions.
Empirical study comparing message passing neural networks and graph transformers for atomistic property prediction.
Theoretical analysis of how attention head count influences transformer approximation properties and expressive power.
Adaptive rollout and routing method for data-driven weather forecasting with improved spatiotemporal modeling.
Novel evaluation metrics for generative models in crystal/material discovery assessing stability, uniqueness, and novelty.
Learning-to-optimize Transformer framework for scalable beamforming in multi-user wireless systems.
Automated algorithm design using machine learning to optimize hyperparameter auto-tuning for high-performance applications.
Geo-Foundation Models framework for flood hazard mapping from SAR satellite imagery in data-scarce regions.
Multi-objective optimization approach for balancing training dynamics across multiple sensor modalities in learning-enabled control systems.
Continual Transformers architecture for real-time low-latency inference on streaming data with reduced redundant computation.
Survey of deep unfolding techniques combining classical optimization algorithms with neural networks for signal processing.
Research on normalization-free transformer architectures using Dynamic Tanh as alternative to standard normalization layers.
Learning-theoretic approach to extracting interpretable features from superposition in complex ML models.
Comparative study of ML methods for forecasting electric vehicle charging demand across different time horizons.
Multimodal VAE method using Hellinger distance and probabilistic opinion pooling for weakly supervised generative learning.
Split learning system using hybrid-order optimization to reduce memory overhead for collaborative LLM training on edge devices.
Architecture separating energy-based world models from language generation in LLMs to improve understanding vs. fluency tradeoff.