4856 papers
This paper challenges the prevailing assumption of Spiking Neural Networks' inherent energy superiority by introducing a rigorous, fair-comparison framework that reveals SNNs only outperform Quantized ANNs under specific low-spike-rate conditions, while demonstrating that such optimized SNNs could nearly double the battery life of devices like smartwatches.
This paper introduces a novel Coupled Oscillator Network (CON) model that overcomes key limitations in latent-space control by ensuring Lagrangian structure, global input-to-state stability, and an invertible input-force mapping, thereby enabling efficient closed-form control strategies for complex mechanical systems using only raw visual feedback.
The paper proposes xTED, a flexible cross-domain adaptation framework that utilizes a diffusion model to edit and transform source domain trajectories into target domain distributions at the data level, thereby bridging domain gaps and enhancing policy learning performance without requiring complex domain-specific modeling.
This paper introduces BNEM, a robust diffusion-based sampler that learns from energy functions via bootstrapped noised energy matching to efficiently generate independent samples from Boltzmann distributions, outperforming existing methods on complex molecular dynamics benchmarks.
This paper establishes that policy gradient methods achieve global convergence with non-asymptotic sample complexity guarantees for finite-horizon MDPs with general state and action spaces by proving the Polyak-Łojasiewicz-Kurdyka condition holds, thereby providing the first theoretical foundations for optimizing multi-period inventory and stochastic cash balance systems.
This paper introduces a constant-lag Neural Delay Differential Equations (NDDEs) framework, inspired by the Mori-Zwanzig formalism, to effectively learn non-Markovian dynamics from partially observed data by identifying memory effects through time delays, demonstrating superior performance over existing methods like LSTMs and ANODEs across synthetic, chaotic, and experimental datasets.
This paper introduces LS-Imagine, a novel approach that enhances open-world reinforcement learning by constructing a long short-term world model with goal-conditioned jumpy transitions and affordance maps, thereby enabling agents to efficiently explore vast state spaces and optimize for long-horizon rewards, as demonstrated by significant improvements in MineDojo.
This paper introduces a PAC-Bayesian framework modeling adversarial training with momentum SGD as a discrete-time dynamical system to derive time-resolved generalization bounds that mechanistically explain robust overfitting and reveal the trade-offs in adversarial weight perturbation design.
This paper demonstrates that frozen transformers, when used in an in-context learning setting, can implicitly infer hidden states to accurately predict the outputs of both linear and nonlinear dynamical systems from noisy observations, achieving performance comparable to optimal and heuristic Bayesian filters without requiring test-time gradient updates or explicit knowledge of the system model.
This paper proposes Adaptive Transfer Clustering (ATC), a unified framework that automatically leverages commonalities between a main and an auxiliary dataset to improve clustering performance despite unknown discrepancies, while providing theoretical optimality guarantees under Gaussian mixture models and demonstrating effectiveness through extensive experiments.
This paper proposes a learned proximal alternating minimization (LPAM) algorithm and its corresponding interpretable network (LPAM-net) for solving two-block nonconvex and nonsmooth optimization problems, proving their convergence to Clarke stationary points and demonstrating superior performance in joint multi-modal MRI reconstruction.
This paper introduces a neurosymbolic system that reconstructs medical images using visual primitives to generate high-level structural explanations, achieving superior classification accuracy and transparency compared to conventional deep learning models in diagnosing histological abnormalities.
The paper introduces Puppet-CNN, a framework that models convolutional layer parameters as states evolving within a learned neural ODE, enabling input-adaptive computation and significant parameter reduction by dynamically determining the effective network depth based on input complexity.
This paper proposes an input-adaptive framework for diffusion models that dynamically adjusts the generative trajectory and sampling steps for each sample based on its complexity, thereby maintaining generation quality while reducing the average number of required steps.
This paper introduces a locomotor task set optimization strategy that uses cluster analysis to identify a minimal, representative subset of tasks for training deep learning models, enabling accurate estimation of hip joint moments for exoskeleton control while significantly reducing data collection requirements.
This paper proposes a parametric, finite-dimensional approach for non-parametric regression that achieves minimax-optimal uniform convergence rates for learning smooth functions and their derivatives while significantly reducing memory and computational costs compared to traditional kernel-based methods.
This paper explores the application of Generative Diffusion Models (GDMs) in massive MIMO communications by reviewing their theoretical foundations, analyzing recent advancements including a case study on near-field channel estimation, and outlining future research directions and challenges for enhancing channel state information acquisition in 5G and 6G networks.
This paper investigates the impact of embedding collapse in Prompt-Tuning by introducing embedding priors, revealing that models can effectively utilize embeddings from diverse activation regions and that distinct activation clusters exist for different task types, suggesting controllable posteriors could enhance interpretability and serve as a foundation for tasks like chain-of-thought distillation.
This paper proposes a method that leverages pretrained vision-language models to learn compact, abstract symbolic world models from limited visual demonstrations, enabling zero-shot generalization and long-horizon planning for complex robotic tasks across novel objects, environments, and goals.
UFGraphFR is a novel federated recommendation framework that enhances privacy-preserving personalization by transforming user data into semantic text vectors to reconstruct global user relationship graphs on the server and employing Transformer architectures for behavior sequence modeling, thereby significantly outperforming existing baselines in accuracy and personalization.