computational efficiency
The capability of an AI algorithm to utilize minimal computational resources (time, memory) while achieving desired performance levels.
- $\texttt{STRCMP}$: Integrating Graph Structural Priors with Language Models for Combinatorial Optimization
- 1000+ FPS 4D Gaussian Splatting for Dynamic Scene Rendering
- A Difference-of-Convex Functions Approach to Energy-Based Iterative Reasoning
- A Unified Framework for Variable Selection in Model-Based Clustering with Missing Not at Random
- AI Progress Should Be Measured by Capability-Per-Resource, Not Scale Alone: A Framework for Gradient-Guided Resource Allocation in LLMs
- AMBER: Adaptive Mesh Generation by Iterative Mesh Resolution Prediction
- Accelerated Vertical Federated Adversarial Learning through Decoupling Layer-Wise Dependencies
- Accelerating 3D Molecule Generative Models with Trajectory Diagnosis
- Accelerating Block Coordinate Descent for LLM Finetuning via Landscape Expansion
- Accelerating Feature Conformal Prediction via Taylor Approximation
- Activation-Informed Merging of Large Language Models
- Adaptive Inference-Time Scaling via Cyclic Diffusion Search
- Adaptive Time Encoding for Irregular Multivariate Time-Series Classification
- Addressing Mark Imbalance in Integration-free Marked Temporal Point Processes
- An Efficient Local Search Approach for Polarized Community Discovery in Signed Networks
- Approximate Domain Unlearning for Vision-Language Models
- Approximately Aligned Decoding
- Attribution-Driven Adaptive Token Pruning for Transformers
- Axial Neural Networks for Dimension-Free Foundation Models
- Balancing Multimodal Training Through Game-Theoretic Regularization
- Better Training Data Attribution via Better Inverse Hessian-Vector Products
- Beyond Greedy Exits: Improved Early Exit Decisions for Risk Control and Reliability
- Beyond Value Functions: Single-Loop Bilevel Optimization under Flatness Conditions
- Bilevel ZOFO: Efficient LLM Fine-Tuning and Meta-Training
- Bio-Inspired Image Restoration
- CDFlow: Building Invertible Layers with Circulant and Diagonal Matrices
- CURE: Concept Unlearning via Orthogonal Representation Editing in Diffusion Models
- Causal Climate Emulation with Bayesian Filtering
- Certifying Deep Network Risks and Individual Predictions with PAC-Bayes Loss via Localized Priors
- Class-aware Domain Knowledge Fusion and Fission for Continual Test-Time Adaptation
- CodeMerge: Codebook-Guided Model Merging for Robust Test-Time Adaptation in Autonomous Driving
- Collapsing Taylor Mode Automatic Differentiation
- Computation and Memory-Efficient Model Compression with Gradient Reweighting
- Computational Algebra with Attention: Transformer Oracles for Border Basis Algorithms
- Computational Budget Should Be Considered in Data Selection
- Computational Efficiency under Covariate Shift in Kernel Ridge Regression
- Conditional Diffusion Anomaly Modeling on Graphs
- Continual Multimodal Contrastive Learning
- Correlation Dimension of Autoregressive Large Language Models
- DC4GS: Directional Consistency-Driven Adaptive Density Control for 3D Gaussian Splatting
- DINO-Foresight: Looking into the Future with DINO
- DSCS: Fast CPDAG-Based Verification of Collapsible Submodels in High-Dimensional Bayesian Networks
- Dendritic Resonate-and-Fire Neuron for Effective and Efficient Long Sequence Modeling
- Depth-Bounds for Neural Networks via the Braid Arrangement
- Derivative-Free Guidance in Continuous and Discrete Diffusion Models with Soft Value-based Decoding
- Diffusion Models Meet Contextual Bandits
- Distilling LLM Agent into Small Models with Retrieval and Code Tools
- Don’t Let It Fade: Preserving Edits in Diffusion Language Models via Token Timestep Allocation
- Dynamic Configuration for Cutting Plane Separators via Reinforcement Learning on Incremental Graph
- Dynamic Focused Masking for Autoregressive Embodied Occupancy Prediction
- Dynamical Low-Rank Compression of Neural Networks with Robustness under Adversarial Attacks
- Dynamical Low-Rank Compression of Neural Networks with Robustness under Adversarial Attacks
- ECO: Evolving Core Knowledge for Efficient Transfer
- ELECTRA: A Cartesian Network for 3D Charge Density Prediction with Floating Orbitals
- Efficient Kernelized Learning in Polyhedral Games beyond Full Information: From Colonel Blotto to Congestion Games
- Efficient Multi-modal Large Language Models via Progressive Consistency Distillation
- Efficient PAC Learning for Realizable-Statistic Models via Convex Surrogates
- Efficient Part-level 3D Object Generation via Dual Volume Packing
- Efficient RAW Image Deblurring with Adaptive Frequency Modulation
- Efficient Safe Meta-Reinforcement Learning: Provable Near-Optimality and Anytime Safety
- Efficient and Near-Optimal Algorithm for Contextual Dueling Bandits with Offline Regression Oracles
- Efficiently Maintaining the Multilingual Capacity of MCLIP in Downstream Cross-Modal Retrieval Tasks
- Energy Loss Functions for Physical Systems
- Enhanced Expert Merging for Mixture-of-Experts in Graph Foundation Models
- Enhancing Training Data Attribution with Representational Optimization
- Enhancing the Outcome Reward-based RL Training of MLLMs with Self-Consistency Sampling
- Explaining Similarity in Vision-Language Encoders with Weighted Banzhaf Interactions
- Exploration via Feature Perturbation in Contextual Bandits
- FADRM: Fast and Accurate Data Residual Matching for Dataset Distillation
- FOCUS: Internal MLLM Representations for Efficient Fine-Grained Visual Question Answering
- FSNet: Feasibility-Seeking Neural Network for Constrained Optimization with Guarantees
- Faithful Group Shapley Value
- FastJAM: a Fast Joint Alignment Model for Images
- Feature-Based Instance Neighbor Discovery: Advanced Stable Test-Time Adaptation in Dynamic World
- FlyLoRA: Boosting Task Decoupling and Parameter Efficiency via Implicit Rank-Wise Mixture-of-Experts
- ForceFM: Enhancing Protein-Ligand Predictions through Force-Guided Flow Matching
- Foresight: Adaptive Layer Reuse for Accelerated and High-Quality Text-to-Video Generation
- Fourier Analysis Network
- FraPPE: Fast and Efficient Preference-Based Pure Exploration
- FreeInv: Free Lunch for Improving DDIM Inversion
- FreqPolicy: Frequency Autoregressive Visuomotor Policy with Continuous Tokens
- Frequency-Aware Token Reduction for Efficient Vision Transformer
- From Information to Generative Exponent: Learning Rate Induces Phase Transitions in SGD
- FuXi-Ocean: A Global Ocean Forecasting System with Sub-Daily Resolution
- FuXi-Ocean: A Global Ocean Forecasting System with Sub-Daily Resolution
- GLVD: Guided Learned Vertex Descent
- GPLQ: A General, Practical, and Lightning QAT Method for Vision Transformers
- Generalized Linear Bandits: Almost Optimal Regret with One-Pass Update
- Generating Full-field Evolution of Physical Dynamics from Irregular Sparse Observations
- Geo-Sign: Hyperbolic Contrastive Regularisation for Geometrically Aware Sign Language Translation
- Geometry Aware Operator Transformer as an efficient and accurate neural surrogate for PDEs on arbitrary domains
- Geometry-Aware Edge Pooling for Graph Neural Networks
- GlobalTomo: A global dataset for physics-ML seismic wavefield modeling and FWI
- GoRA: Gradient-driven Adaptive Low Rank Adaptation
- High-order Equivariant Flow Matching for Density Functional Theory Hamiltonian Prediction
- Hippocampal-like Sequential Editing for Continual Knowledge Updates in Large Language Models
- How Many Tokens Do 3D Point Cloud Transformer Architectures Really Need?
- How Well Can Differential Privacy Be Audited in One Run?
- Hyperbolic Dataset Distillation
- IF-Guide: Influence Function-Guided Detoxification of LLMs
- Implicit-ARAP: Efficient Handle-Guided Neural Field Deformation via Local Patch Meshing
- Improved Algorithms for Overlapping and Robust Clustering of Edge-Colored Hypergraphs: An LP-Based Combinatorial Approach
- Infrequent Exploration in Linear Bandits
- Is Noise Conditioning Necessary? A Unified Theory of Unconditional Graph Diffusion Models
- JAMUN: Bridging Smoothed Molecular Dynamics and Score-Based Learning for Conformational Ensemble Generation
- JanusDNA: A Powerful Bi-directional Hybrid DNA Foundation Model
- KAIROS: Scalable Model-Agnostic Data Valuation
- Kinaema: a recurrent sequence model for memory and pose in motion
- LIMOPro: Reasoning Refinement for Efficient and Effective Test-time Scaling
- LLM at Network Edge: A Layer-wise Efficient Federated Fine-tuning Approach
- Learning Gradient Boosted Decision Trees with Algorithmic Recourse
- Learning Latent Variable Models via Jarzynski-adjusted Langevin Algorithm
- Learning with Calibration: Exploring Test-Time Computing of Spatio-Temporal Forecasting
- Less Is More, but Where? Dynamic Token Compression via LLM-Guided Keyframe Prior
- LoRASuite: Efficient LoRA Adaptation Across Large Language Model Upgrades
- LoTA-QAF: Lossless Ternary Adaptation for Quantization-Aware Fine-Tuning
- Low-Rank Graphon Learning for Networks
- MAT-Agent: Adaptive Multi-Agent Training Optimization
- ML4CFD Competition: Results and Retrospective Analysis
- MaNGO — Adaptable Graph Network Simulators via Meta-Learning
- Mask Image Watermarking
- Memory-Augmented Potential Field Theory: A Framework for Adaptive Control in Non-Convex Domains
- Memory-Efficient Training with In-Place FFT Implementation
- Metropolis Adjusted Microcanonical Hamiltonian Monte Carlo
- Mind the Gap: Removing the Discretization Gap in Differentiable Logic Gate Networks
- MoESD: Unveil Speculative Decoding's Potential for Accelerating Sparse MoE
- More Than Just Functional: LLM-as-a-Critique for Efficient Code Generation
- Multi-Agent Collaboration via Evolving Orchestration
- Multi-View Oriented GPLVM: Expressiveness and Efficiency
- Neural Green’s Functions
- Noise Consistency Training: A Native Approach for One-step Generator in Learning Additional Controls
- Non-exchangeable Conformal Prediction with Optimal Transport: Tackling Distribution Shift with Unlabeled Data
- On Agnostic PAC Learning in the Small Error Regime
- On Efficiency-Effectiveness Trade-off of Diffusion-based Recommenders
- One Head to Rule Them All: Amplifying LVLM Safety through a Single Critical Attention Head
- One Sample is Enough to Make Conformal Prediction Robust
- Online Locally Differentially Private Conformal Prediction via Binary Inquiries
- Orochi: Versatile Biomedical Image Processor
- Over-squashing in Spatiotemporal Graph Neural Networks
- PAC-Bayes Bounds for Multivariate Linear Regression and Linear Autoencoders
- PARCO: Parallel AutoRegressive Models for Multi-Agent Combinatorial Optimization
- PPMStereo: Pick-and-Play Memory Construction for Consistent Dynamic Stereo Matching
- Parallel Scaling Law for Language Models
- Partial Correlation Network Estimation by Semismooth Newton Methods
- Partial Information Decomposition via Normalizing Flows in Latent Gaussian Distributions
- Practical Kernel Selection for Kernel-based Conditional Independence Test
- Pre-trained Large Language Models Learn to Predict Hidden Markov Models In-context
- Privacy amplification by random allocation
- Provably Efficient Online RLHF with One-Pass Reward Modeling
- Provably Efficient RL under Episode-Wise Safety in Constrained MDPs with Linear Function Approximation
- REP: Resource-Efficient Prompting for Rehearsal-Free Continual Learning
- RaySt3R: Predicting Novel Depth Maps for Zero-Shot Object Completion
- Reinforcement Learning for Out-of-Distribution Reasoning in LLMs: An Empirical Study on Diagnosis-Related Group Coding
- Return of ChebNet: Understanding and Improving an Overlooked GNN on Long Range Tasks
- Reverse-Annealed Sequential Monte Carlo for Efficient Bayesian Optimal Experiment Design
- Revisiting End-to-End Learning with Slide-level Supervision in Computational Pathology
- Revisiting Frank-Wolfe for Structured Nonconvex Optimization
- RobIA: Robust Instance-aware Continual Test-time Adaptation for Deep Stereo
- Robust Regression of General ReLUs with Queries
- SANSA: Unleashing the Hidden Semantics in SAM2 for Few-Shot Segmentation
- SHF: Symmetrical Hierarchical Forest with Pretrained Vision Transformer Encoder for High-Resolution Medical Segmentation
- SHGR: A Generalized Maximal Correlation Coefficient
- Safely Learning Controlled Stochastic Dynamics
- Sample-Conditional Coverage in Split-Conformal Prediction
- Sampled Estimators For Softmax Must Be Biased
- Sampling 3D Molecular Conformers with Diffusion Transformers
- Scaling Epidemic Inference on Contact Networks: Theory and Algorithms
- Score-informed Neural Operator for Enhancing Ordering-based Causal Discovery
- Simple and Effective Specialized Representations for Fair Classifiers
- SparseDiT: Token Sparsification for Efficient Diffusion Transformer
- Spike-RetinexFormer: Rethinking Low-light Image Enhancement with Spiking Neural Networks
- Spiking Meets Attention: Efficient Remote Sensing Image Super-Resolution with Attention Spiking Neural Networks
- Split conformal classification with unsupervised calibration
- Stable Coresets via Posterior Sampling: Aligning Induced and Full Loss Landscapes
- Statistics Caching Test-Time Adaptation for Vision-Language Models
- Styl3R: Instant 3D Stylized Reconstruction for Arbitrary Scenes and Styles
- Synergy Between the Strong and the Weak: Spiking Neural Networks are Inherently Self-Distillers
- TEMPO: Temporal Multi-scale Autoregressive Generation of Protein Conformational Ensembles
- TF-MAS: Training-free Mamba2 Architecture Search
- Temporal Chain of Thought: Long-Video Understanding by Thinking in Frames
- The Adaptive Complexity of Minimizing Relative Fisher Information
- The Omni-Expert: A Computationally Efficient Approach to Achieve a Mixture of Experts in a Single Expert Model
- Time-uniform and Asymptotic Confidence Sequence of Quantile under Local Differential Privacy
- To Distill or Decide? Understanding the Algorithmic Trade-off in Partially Observable RL
- Towards a Golden Classifier-Free Guidance Path via Foresight Fixed Point Iterations
- Towards foundational LiDAR world models with efficient latent flow matching
- Training-Free Efficient Video Generation via Dynamic Token Carving
- Training-Free Test-Time Adaptation via Shape and Style Guidance for Vision-Language Models
- TrajMamba: An Efficient and Semantic-rich Vehicle Trajectory Pre-training Model
- Transformers Learn Faster with Semantic Focus
- UFO-RL: Uncertainty-Focused Optimization for Efficient Reinforcement Learning Data Selection
- UGM2N: An Unsupervised and Generalizable Mesh Movement Network via M-Uniform Loss
- Ultra-high Resolution Watermarking Framework Resistant to Extreme Cropping and Scaling
- Ultrametric Cluster Hierarchies: I Want ‘em All!
- Uncertainty Estimation by Flexible Evidential Deep Learning
- Uncertainty Quantification with the Empirical Neural Tangent Kernel
- VITRIX-UniViTAR: Unified Vision Transformer with Native Resolution
- VORTA: Efficient Video Diffusion via Routing Sparse Attention
- Venus-MAXWELL: Efficient Learning of Protein-Mutation Stability Landscapes using Protein Language Models
- When Do Transformers Outperform Feedforward and Recurrent Networks? A Statistical Perspective
- YOLOv12: Attention-Centric Real-Time Object Detectors
- ZigzagPointMamba: Spatial-Semantic Mamba for Point Cloud Understanding