inference optimization

Dynamic-Width Speculative Beam Decoding for LLM Inference AAAI 2025 ECHO-LLaMA: Efficient Caching for High-Performance LLaMA Training EMNLP 2025 EcoTune: Token-Efficient Multi-Fidelity Hyperparameter Optimization for Large Language Model Inference EMNLP 2025 Efficiently Editing Mixture-of-Experts Models with Compressed Experts EMNLP 2025 Energy Considerations of Large Language Model Inference and Efficiency Optimizations ACL 2025 Exploring Hybrid Sampling Inference for Aspect-based Sentiment Analysis NAACL 2025 Faster Cascades via Speculative Decoding ICLR 2025 Fast Quiet-STaR: Thinking Without Thought Tokens EMNLP 2025 FBQuant: FeedBack Quantization for Large Language Models IJCAI 2025 Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success RSS 2025 Flaming-hot Initiation with Regular Execution Sampling for Large Language Models NAACL 2025 Flash Inference: Near Linear Time Inference for Long Convolution Sequence Models and Beyond ICLR 2025 From Implicit Exploration to Structured Reasoning: Guideline and Refinement for LLMs EMNLP 2025 Fuzzy Speculative Decoding for a Tunable Accuracy-Runtime Tradeoff ACL 2025 GraphKV: Breaking the Static Selection Paradigm with Graph-Based KV Cache Eviction EMNLP 2025 Hardware-Aware Parallel Prompt Decoding for Memory-Efficient Acceleration of LLM Inference EMNLP 2025 HARP: Hesitation-Aware Reframing in Transformer Inference Pass NAACL 2025 Inference Compute-Optimal Video Vision Language Models ACL 2025 Inference Optimal VLMs Need Fewer Visual Tokens and More Parameters ICLR 2025 Inference-Time Diffusion Model Distillation ICCV 2025 LAMB: A Training-Free Method to Enhance the Long-Context Understanding of SSMs via Attention-Guided Token Filtering ACL 2025 Let the Code LLM Edit Itself When You Edit the Code ICLR 2025 Libra-Merging: Importance-redundancy and Pruning-merging Trade-off for Acceleration Plug-in in Large Vision-Language Model CVPR 2025 Lookahead Q-Cache: Achieving More Consistent KV Cache Eviction via Pseudo Query EMNLP 2025 MiniKV: Pushing the Limits of 2-Bit KV Cache via Compression and System Co-Design for Efficient Long Context Inference ACL 2025