Transformer终结者？Google DeepMind新架构实现2倍推理速度和一半内存占用
"计算效率" 相关文章
TransPrune: Token Transition Pruning for Efficient Large Vision-Language Model
Efficient Knowledge Tracing Leveraging Higher-Order Information in Integrated Graphs
注意力机制洗牌！GTA硬刚MHA/GQA：计算减半不减分，缓存压缩还提速
English writes numbers backwards
Neuro-MSBG: An End-to-End Neural Model for Hearing Loss Simulation
APTx Neuron: A Unified Trainable Neuron Architecture Integrating Activation and Computation
A Sparsity Predicting Approach for Large Language Models via Activation Pattern Clustering
Learning Null Geodesics for Gravitational Lensing Rendering in General Relativity
ProofCompass: Enhancing Specialized Provers with LLM Guidance
「有望成为Transformer杀手」，谷歌DeepMind新架构MoR实现两倍推理速度
「有望成为Transformer杀手」，谷歌DeepMind新架构MoR实现两倍推理速度
Fragment size density estimator for shrinkage-induced fracture based on a physics-informed neural network
FactorHD: A Hyperdimensional Computing Model for Multi-Object Multi-Class Representation and Factorization
Mixture of Raytraced Experts
Scalable Unsupervised Segmentation via Random Fourier Feature-based Gaussian Process
Mixture of Experts in Large Language Models
LoRA Is Slower Than You Think
[问与答] 咨询一个算法的问题，我是一点思路都没有了。
[问与答] 咨询一个算法的问题，我是一点思路都没有了。
