All papers

This paper introduces a reinforcement learning navigation policy using shared spatial fields for large-scale lifelong multi-agent path finding.

This paper proposes a cross-image routing attention mechanism to predict millimeter-wave beam power from dual-endpoint vehicle-to-infrastructure camera views.

This paper presents an operator-valued kernel regression framework that enforces aggregation constraints directly in hierarchical time series forecasting.

This book chapter provides an overview and mathematical introduction to finite-parametric operator surrogates for approximating maps between infinite-dimensional spaces.

This paper introduces a spatial transformer opponent model that stabilizes reinforcement learning training under partial observability in competitive foraging games.

This paper proposes Lookahead Self-Distillation, providing cross-instance token-level supervision to improve reinforcement learning with verifiable rewards for language models.

This paper develops an online algorithm achieving optimal regret bounds for stochastic contextual bandits with unknown context-lumpable cluster structures.

This paper introduces a physics-informed graph neural network for modeling advection-diffusion processes to forecast solar power production.

This paper derives finite-sample risk bounds for Kernel Mean Matching under dependent sampling using effective sample sizes.

This paper presents HE-CV-UCB, an algorithm for Bayesian optimization of Conditional Value-at-Risk under unknown Gaussian process hyperparameters.

This paper analyzes the geometric robustness of over-parameterized low-rank matrix sensing under the Burer-Monteiro factorization.

This paper provides optimal learning rate bounds for random feature kernel ridge regression under misspecified models and heavy-tailed noise.

This paper develops a theoretical framework characterizing when representations learned on a source task remain recoverable and reusable for future tasks.

This paper proposes a forecast-error distribution entropy framework to detect nonlinear predictive coupling across multiple time horizons.

This paper introduces LunaTP, a parallel execution framework that accelerates tensor product convolutions in equivariant machine learning interatomic potentials.

This paper benchmarks eighteen continual learning methods on a sequential radar waveform recognition task across task- and class-incremental settings.

This paper develops finite-sample bounds to guide algorithm design for Expected Improvement in continuous simulation optimization with heteroscedastic noise.

This paper presents FastEEP, a system that speeds up intra-instance elastic scaling for Mixture-of-Experts serving by reusing runtime state.

This paper develops graph metanetworks that respect the weight-space symmetries of transformer language models, including attention matrices and token embeddings.

This paper introduces ProTok, a unified discrete tokenizer for protein sequence, structure, and functional annotations using modular multi-codebook quantization.

This paper presents Lifting, a framework combining repetition encoding and neural decoding for mid-circuit quantum state discrimination on superconducting hardware.

This paper provides a systematization of knowledge on membership inference attacks in tabular machine learning across small-to-medium datasets.

This paper provides minimax rates and limits for recalibrating predictions across many binary questions when each question has few ground-truth labels.

This paper introduces a risk-seeking optimization method to prepare target quantum states with high fidelity on noisy bosonic quantum hardware.

This paper proposes LlMBO, applying Bayesian optimization to ultra-low-dimensional fine-tuning as an alternative to gradient-based policy optimization.

This paper analyzes heterogeneous asynchronous multi-player bandits, proves the centralized allocation problem is NP-hard, and studies greedy approximation policies.

This paper introduces CACTUS, an ADMM-based optimizer that speeds up curvature-aware pruning and preserves influence functions in compressed language models.

This paper replaces squared Euclidean distance with Euclidean distance in Flow Matching to improve distributional robustness against data contamination.

The paper derives explicit variance approximations to extend circulant permutation techniques to different MinHash estimators.

The paper provides last-iterate convergence bounds for clipped stochastic gradient descent under heavy-tailed noise and causal adversarial perturbations.

The paper establishes linear last-iterate convergence rates for optimistic learning algorithms in multiplayer harmonic games.

The paper analyzes how prediction errors affect demographic parity fairness violations in budget-constrained allocation problems.

The paper proposes a flow-matching model that integrates deterministic forecasting as a velocity component for stable probabilistic time-series predictions.

The paper derives minimax risk bounds showing how covariate design smoothness affects pointwise conditional average treatment effect estimation.

The paper presents a profiling-guided framework to allocate precision budgets adaptively across layers and tokens during KV cache quantization.

The paper introduces a training-free conditional sampling method for pretrained Schrödinger bridge generative models.

The paper demonstrates that the placement timing of curvature updates in Kronecker-preconditioned optimizers significantly affects language model training loss.

The paper proves theoretical teaching complexity bounds for closed neighborhoods in planar graphs without collusion.

The paper compares transformer sequence predictions against an optimal Bayesian predictor across structured Markov sources.

The paper proposes a reinforcement learning method to selectively transfer data for carbon-footprint modeling in new manufacturing settings.

The paper presents an adaptive, training-free speculative decoding method to speed up inference for vision-language models.

The paper develops an off-policy reinforcement learning method with token-wise importance sampling and proximal KL regularization.

The paper analyzes how network depth influences the internal loop geometry and spatial reach of gauge-equivariant graph neural networks.

The paper formalizes layer-to-layer representation sharing as a memory storage and retrieval framework for deep transformer networks.

The paper analytically characterizes the mechanisms and conditions behind the grokking phenomenon in a simplified learning setup.

The paper presents a neural network architecture for simultaneous model selection and parameter posterior estimation in chemical reaction networks.

The paper introduces methods to generate graphs whose topological persistence diagrams match target descriptors exactly rather than approximately.

The paper presents a method for training diffusion models in lattice field theory directly from their own generated samples.

The paper analyzes training instability cascades in large recommender systems and proposes targeted gradient, weight, and loss interventions.

The paper introduces a framework that uses language models to extract task-scheduling priority functions from optimal solver solutions on directed acyclic graphs.

The paper proposes a continuous probabilistic Gaussian mixture framework to improve human pose estimation under dense occlusions without retraining models.

This paper evaluates category name matching versus pairwise geometry metrics for candidate retrieval in context-constrained 3D vision-language models.

This paper evaluates calibration differences between object-level softmax and agreement metrics when resolving ambiguous 3D referring queries.

This paper presents a flow-matching transformer pipeline to generate full gigapixel histopathology whole-slide images.

This paper studies transformer architectural choices for handling missing modalities and longitudinal observations in multimodal healthcare datasets.

This paper presents a unified model for dense visual correspondence across 25 pairs of multi-modal sensor inputs.

The paper proposes a reinforcement learning framework using natural language metrics to reduce object hallucination in vision-language image captioning.

The paper presents a four-gate verification protocol to evaluate whether audio-conditioned visual prediction models genuinely rely on input audio signals.

The paper introduces a lightweight test-time attention framework that leverages single-photon sensor noise as an informative prior for indoor depth estimation.

This paper introduces prototype-purity pseudo-labeling to extract accurate instance-level annotations from weakly supervised whole-slide pathology images.