Optimization & Theory

HeSS: Head Sensitivity Score for Sparsity Redistribution in VGGT CVPR 2026 Heterogeneity at Hyperscale: Characterization and Scheduling of Large Production AI Clusters at Alibaba (Operational Systems) OSDI 2026 HeteroSpec: Leveraging Contextual Heterogeneity for Efficient Speculative Decoding ACL 2026 Hetu v2: A General and Scalable Deep Learning System with Hierarchical and Heterogeneous Single Program Multiple Data Annotations OSDI 2026 Hidden Monotonicity: Explaining Deep Neural Networks via their DC Decomposition CVPR 2026 High-Dimensional Analysis of Gradient Flow for Extensive-Width Quadratic Neural Networks JMLR 2026 Hi-Lo Prune: Look at What You'll Lose before Pruning with Hierarchical Token Selection CVPR 2026 HiSVD: Principled Low-Rank Approximation of LLMs via Hierarchical Modeling of Information Capacity and Spectral Structure ACL 2026 How Does the ReLU Activation Affect the Implicit Bias of Gradient Descent on High-dimensional Neural Network Regression? COLT 2026 How Instruction and Reasoning Data shape Post-Training: Data Quality through the Lens of Layer-wise Gradients ACL 2026 How to Set β_1, β_2 in Adam: An Online Learning Perspective ALT 2026 H-Sets: Hessian-Guided Discovery of Set-Level Feature Interactions in Image Classifiers CVPR 2026 HTTM: Head-wise Temporal Token Merging for Faster VGGT CVPR 2026 HybridKV: Hybrid KV Cache Compression for Efficient Multimodal Large Language Model Inference ACL 2026 Hybrid Token Compression for Vision-Language Models CVPR 2026 Hyperion: Private Token Sampling with Homomorphic Encryption ACL 2026 HyperNAS: Enhancing Architecture Representation for NAS Predictor via Hypernetwork CVPR 2026 I2I-Bench: A Comprehensive Benchmark Suite for Image-to-Image Editing Models CVPR 2026 IAFMNet: Information-Aware Feature Modulation for Efficient Super-Resolution CVPR 2026 IF-Prune: Information-Flow Guided Token Pruning for Efficient Vision-Language Models CVPR 2026 Image Diffusion Preview with Consistency Solver CVPR 2026 IMPACT: Importance-Aware Activation Space Reconstruction ACL 2026 Improving Calibration in Test-Time Prompt Tuning for Vision-Language Models via Data-Free Flatness-Aware Prompt Pretraining CVPR 2026 Improving Diffusion Generalization with Weak-to-Strong Segmented Guidance CVPR 2026 Inside-Out: Measuring Generalization in Vision Transformers Through Inner Workings CVPR 2026