296 papers
This paper proposes "amortized MIPS," a learning-based framework that leverages the mathematical properties of support functions to train neural networks (SupportNet and KeyNet) that directly predict optimal keys for Maximum Inner Product Search, thereby amortizing computational costs for queries drawn from a fixed distribution.
This paper proposes a Bayesian Transformer framework that integrates Monte Carlo Dropout, variational feed-forward layers, and stochastic attention into a PatchTST backbone to deliver well-calibrated, state-of-the-art probabilistic load forecasts with robust uncertainty estimates under extreme weather distributional shifts across multiple global power grids.
This paper introduces a particle filtering framework to rigorously analyze the accuracy-cost tradeoffs of parallel inference methods in large language models, establishing theoretical guarantees and identifying fundamental limits while demonstrating that sampling error alone does not fully predict final model accuracy.
This paper establishes finite-sample guarantees for cost-driven state representation learning in infinite-horizon time-invariant Linear Quadratic Gaussian (LQG) control by analyzing two approaches—explicit latent modeling and implicit MuZero-like dynamics—while introducing a key technical proof of persistency of excitation for a novel stochastic process arising from quadratic regression.
This paper introduces an adversarial latent-initial-state POMDP framework that theoretically establishes a minimax principle and finite-sample guarantees, while empirically demonstrating that targeted adversarial training significantly reduces robustness gaps in partially observable reinforcement learning.
This paper introduces MSKernelBench, a comprehensive benchmark covering diverse multi-scenario GPU kernels, and CUDAMaster, a multi-agent, hardware-aware system that leverages this benchmark to achieve significant speedups, often matching or surpassing closed-source libraries like cuBLAS, thereby advancing general-purpose automated CUDA kernel optimization beyond current ML-focused methods.
This paper proposes DualAdam, a novel optimizer that combines the standard Adam algorithm with its inverse variant, InvAdam, to mathematically demonstrate and empirically validate improved generalization by helping models escape sharp minima and converge to flatter ones.
This paper introduces the Combinatorial Allocation Bandits (CAB) framework to optimize arm satisfaction in matching platforms rather than just maximizing match counts, proposing and analyzing both Upper Confidence Bound and Thompson Sampling algorithms that achieve near-optimal regret bounds under a generalized linear model.
NEST is a novel device placement framework that unifies network, compute, and memory awareness through structured dynamic programming to jointly optimize hybrid parallelism strategies, achieving up to 2.43 times higher throughput and improved scalability compared to state-of-the-art baselines.
This paper proposes the Latent Autoencoder Ensemble Kalman Filter (LAE-EnKF), a novel data assimilation method that learns a stable, linear state-space model in a latent space to overcome the performance limitations of standard EnKF on strongly nonlinear and chaotic systems while maintaining computational efficiency.
This paper introduces the Khatri-Rao clustering paradigm, which extends traditional centroid-based methods like k-Means and deep clustering by modeling centroids as interactions of multiple succinct protocentroids, thereby achieving more compact and accurate data summaries with reduced redundancy.
The paper introduces HaKAN, a lightweight and interpretable time series forecasting model that leverages Hahn polynomial-based Kolmogorov-Arnold Networks (KANs) with channel independence and patching to effectively capture both global and local temporal patterns, outperforming recent state-of-the-art Transformer and MLP-based methods.
This paper proposes a novel sampling method for unnormalized Boltzmann densities that leverages a sequence of Langevin samplers to efficiently simulate a probability flow ODE derived from linear stochastic interpolants by generating intermediate samples and robustly estimating the velocity field, while providing theoretical convergence guarantees and demonstrating effectiveness on challenging multimodal distributions and Bayesian inference tasks.
This paper introduces "Bayesian wind tunnels" to rigorously demonstrate that small transformers perform exact Bayesian inference through a specific geometric mechanism involving residual streams as belief substrates and attention-based routing, a capability that capacity-matched MLPs fundamentally lack.
This paper introduces novel absolute cluster validity indices that quantify the compactness and separability of clusters to determine the true number of clusters, demonstrating their effectiveness across synthetic and real-world datasets compared to existing relative indices.
This paper proposes a Cholesky-based algorithm for Covariance-Adjusted Support Vector Classification that overcomes the sub-optimality of traditional max-margin KKT conditions in non-Euclidean spaces by incorporating class covariance structures, resulting in significantly improved classification performance across multiple datasets.
This paper proposes an optimal transport-based aggregation framework that efficiently combines locally trained Mixture-of-Experts models into a global estimator with a single communication step, achieving performance comparable to centralized training while significantly reducing computational and communication costs.
This paper introduces a disjunctive branch-and-bound framework combined with novel convex relaxations to solve low-rank matrix completion problems to certifiable optimality, significantly reducing optimality and test errors compared to existing heuristic methods for matrices up to 2500 dimensions.
This paper introduces a unified framework for online LLM watermark detection based on e-processes, which provides anytime-valid statistical guarantees and enhances detection power through empirically adaptive methods applicable to various sequential testing problems.
This paper proposes a fast, deterministic tensor completion method that leverages standard linear algebra to recover tensors from fiberwise observations along a single mode, offering efficient recovery guarantees without relying on random sampling assumptions.