performance optimization
The process of improving the efficiency and effectiveness of an AI model, often through techniques such as hyperparameter tuning, model selection, or algorithm refinement, to achieve better results on specific tasks.
- AI Research Agents for Machine Learning: Search, Exploration, and Generalization in MLE-bench
- AltLoRA: Towards Better Gradient Approximation in Low-Rank Adaptation with Alternating Projections
- Automated Composition of Agents: A Knapsack Approach for Agentic Component Selection
- Combining Cost Constrained Runtime Monitors for AI Safety
- Critical Batch Size Revisited: A Simple Empirical Approach to Large-Batch Language Model Training
- CyIN: Cyclic Informative Latent Space for Bridging Complete and Incomplete Multimodal Learning
- Data Mixture Optimization: A Multi-fidelity Multi-scale Bayesian Framework
- Datasets, Documents, and Repetitions: The Practicalities of Unequal Data Quality
- Decoupled Entropy Minimization
- Distance Adaptive Beam Search for Provably Accurate Graph-Based Nearest Neighbor Search
- EraseFlow: Learning Concept Erasure Policies via GFlowNet-Driven Alignment
- Fast Training of Large Kernel Models with Delayed Projections
- Fuse2Match: Training-Free Fusion of Flow, Diffusion, and Contrastive Models for Zero-Shot Semantic Matching
- Gated Integration of Low-Rank Adaptation for Continual Learning of Large Language Models
- Generating and Checking DNN Verification Proofs
- Improving Data Efficiency for LLM Reinforcement Fine-tuning Through Difficulty-targeted Online Data Selection and Rollout Replay
- Improving Energy Natural Gradient Descent through Woodbury, Momentum, and Randomization
- Incentivizing Time-Aware Fairness in Data Sharing
- Inference-Time Scaling for Flow Models via Stochastic Generation and Rollover Budget Forcing
- Learning to Learn with Contrastive Meta-Objective
- Learning to Learn with Contrastive Meta-Objective
- Length Generalization via Auxiliary Tasks
- MIDAS: Misalignment-based Data Augmentation Strategy for Imbalanced Multimodal Learning
- Multi-Agent Collaboration via Evolving Orchestration
- NaViL: Rethinking Scaling Properties of Native Multimodal Large Language Models under Data Constraints
- On Vanishing Gradients, Over-Smoothing, and Over-Squashing in GNNs: Bridging Recurrent and Graph Learning
- Optimal Dynamic Regret by Transformers for Non-Stationary Reinforcement Learning
- RTV-Bench: Benchmarking MLLM Continuous Perception, Understanding and Reasoning through Real-Time Video
- Rethinking Fine-Tuning when Scaling Test-Time Compute: Limiting Confidence Improves Mathematical Reasoning
- Retrv-R1: A Reasoning-Driven MLLM Framework for Universal and Efficient Multimodal Retrieval
- S$^2$NN: Sub-bit Spiking Neural Networks
- Table2LaTeX-RL: High-Fidelity LaTeX Code Generation from Table Images via Reinforced Multimodal Language Models
- Tensor-Parallelism with Partially Synchronized Activations
- The Rise of Parameter Specialization for Knowledge Storage in Large Language Models
- Towards Thinking-Optimal Scaling of Test-Time Compute for LLM Reasoning
- What Moves the Eyes: Doubling Mechanistic Model Performance Using Deep Networks to Discover and Test Cognitive Hypotheses
- Why Masking Diffusion Works: Condition on the Jump Schedule for Improved Discrete Diffusion
- Win Fast or Lose Slow: Balancing Speed and Accuracy in Latency-Sensitive Decisions of LLMs