Optimization

TREAD: Token Routing for Efficient Architecture-agnostic Diffusion Training ICCV 2025 TurboRAG: Accelerating Retrieval-Augmented Generation with Precomputed KV Caches for Chunked Text EMNLP 2025 Understanding Flatness in Generative Models: Its Role and Benefits ICCV 2025 Understanding Silent Data Corruption in LLM Training ACL 2025 Unifying Uniform and Binary-coding Quantization for Accurate Compression of Large Language Models ACL 2025 Unleashing the Potential of Large Language Models as Prompt Optimizers: Analogical Analysis with Gradient-based Model Optimizers AAAI 2025 UP-Restorer: When Unrolling Meets Prompts for Unified Image Restoration AAAI 2025 Using Powerful Prior Knowledge of Diffusion Model in Deep Unfolding Networks for Image Compressive Sensing CVPR 2025 Velocitune: A Velocity-based Dynamic Domain Reweighting Method for Continual Pre-training ACL 2025 VLMInferSlow: Evaluating the Efficiency Robustness of Large Vision-Language Models as a Service ACL 2025 Wasserstein Distance Constraint and Parameter Sparsification for Batched and Iterative Knowledge Editing AAAI 2025 Why Are Positional Encodings Nonessential for Deep Autoregressive Transformers? A Petroglyph Revisited ACL 2025 WINS: Winograd Structured Pruning for Fast Winograd Convolution ICCV 2025 YuLan-Mini: Pushing the Limits of Open Data-efficient Language Model ACL 2025 Z1: Efficient Test-time Scaling with Code EMNLP 2025 ZEN: Empowering Distributed Training with Sparsity-driven Data Synchronization OSDI 2025 $\boldsymbol{\mu}\mathbf{P^2}$: Effective Sharpness Aware Minimization Requires Layerwise Perturbation Scaling NeurIPS 2024 2DQuant: Low-bit Post-Training Quantization for Image Super-Resolution NeurIPS 2024 Accelerating Adversarially Robust Model Selection for Deep Neural Networks via Racing AAAI 2024 Accelerating Diffusion Sampling with Optimized Time Steps CVPR 2024 Accelerating Greedy Coordinate Gradient and General Prompt Optimization via Probe Sampling NeurIPS 2024 Accelerating Relative Entropy Coding with Space Partitioning NeurIPS 2024 Adam with model exponential moving average is effective for nonconvex optimization NeurIPS 2024 Adaptive Multi-Modal Cross-Entropy Loss for Stereo Matching CVPR 2024 Adaptive Time-Stepping Schedules for Diffusion Models UAI 2024