Two-Stage Surrogate Modeling for Data-Driven Design Optimization with Application to Composite Microstructure Generationby Farhad Pourkamali-Anaraki,…
Theoretical guarantees on the best-of-n alignment policyby Ahmad Beirami, Alekh Agarwal, Jonathan Berant, Alexander D'Amour,…
Shrinking Your TimeStep: Towards Low-Latency Neuromorphic Object Recognition with Spiking Neural Networkby Yongqi Ding, Lin…
LLM Maybe LongLM: Self-Extend LLM Context Window Without Tuningby Hongye Jin, Xiaotian Han, Jingfeng Yang,…
TPC-ViT: Token Propagation Controller for Efficient Vision Transformerby Wentao ZhuFirst submitted to arxiv on: 3…
Unsupervised Outlier Detection using Random Subspace and Subsampling Ensembles of Dirichlet Process Mixturesby Dongwook Kim,…
SecFormer: Fast and Accurate Privacy-Preserving Inference for Transformer Models via SMPCby Jinglong Luo, Yehong Zhang,…
Real-Time FJ/MAC PDE Solvers via Tensorized, Back-Propagation-Free Optical PINN Trainingby Yequan Zhao, Xian Xiao, Xinling…
Energy-Efficient Power Control for Multiple-Task Split Inference in UAVs: A Tiny Learning-Based Approachby Chenxi Zhao,…
Beyond Chinchilla-Optimal: Accounting for Inference in Language Model Scaling Lawsby Nikhil Sardana, Jacob Portes, Sasha…