Optimization

Tessera: A Holistic Pipeline Parallelism Framework for Trillion-Parameter Heterogeneous MoE Training (Operational Systems) OSDI 2026 Test-Time Alignment of Text-to-Image Diffusion Models via Null-Text Embedding Optimisation CVPR 2026 The Convergence Behavior of Adam under Heavy-Tailed Noise UAI 2026 The Devil Is in Gradient Entanglement: Energy-Aware Gradient Coordinator for Robust Generalized Category Discovery CVPR 2026 The Power of Decaying Steps: Enhancing Attack Stability and Transferability for Sign-based Optimizers CVPR 2026 UniComp: Rethinking Video Compression Through Informational Uniqueness CVPR 2026 UniFusion: A Unified Image Fusion Framework with Robust Representation and Source-Aware Preservation CVPR 2026 Upsample Anything: A Simple and Hard to Beat Baseline for Feature Upsampling CVPR 2026 Vanilla SGD with Momentum Survives Heavy-Tailed Noise: Convergence Analysis without Gradient Clipping or Normalization UAI 2026 VCORE: Variance-Controlled Optimization-based Reweighting for Chain-of-Thought Supervision ACL 2026 VisPCO: Visual Token Pruning Configuration Optimization via Budget-Aware Pareto-Frontier Learning for Vision-Language Models ACL 2026 When Both Layers Learn: Training Dynamics of Representing Linear Models via ReLU Networks COLT 2026 Why Not Hyperparameter-Friendly Optimisation? A Monotonic Adaptive Norm Rescaling Approach For Long-Tailed Recognition CVPR 2026 WildPose: A Unified Framework for Robust Pose Estimation in the Wild CVPR 2026 Accelerated Test-Time Scaling with Model-Free Speculative Sampling EMNLP 2025 Accelerate Parallelizable Reasoning via Parallel Decoding within One Sequence EMNLP 2025 Accelerating Adversarial Training on Under-Utilized GPU IJCAI 2025 Accelerating Diffusion Transformer via Increment-Calibrated Caching with Channel-Aware Singular Value Decomposition CVPR 2025 A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training CVPR 2025 AdaDPCC: Adaptive Rate Control and Rate-Distortion-Complexity Optimization for Dynamic Point Cloud Compression AAAI 2025 AdamS: Momentum Itself Can Be A Normalizer for LLM Pretraining and Post-training EMNLP 2025 Adaptive backtracking line search ICLR 2025 Adaptive Draft-Verification for Efficient Large Language Model Decoding AAAI 2025 Adaptive Dual-domain Learning for Underwater Image Enhancement AAAI 2025 Adaptive Guidance: Training-free Acceleration of Conditional Diffusion Models AAAI 2025