Efficiera Residual Networks: Hardware-Friendly Fully Binary Weight with 2-bit Activation Model Achieves Practical ImageNet Accuracyby…
Data Quality Control in Federated Instruction-tuning of Large Language Modelsby Yaxin Du, Rui Ye, Fengting…
PaSTe: Improving the Efficiency of Visual Anomaly Detection at the Edgeby Manuel Barusco, Francesco Borsatti,…
Enhancing Unimodal Latent Representations in Multimodal VAEs through Iterative Amortized Inferenceby Yuta Oshima, Masahiro Suzuki,…
A Theoretical Survey on Foundation Modelsby Shi Fu, Yuzhu Chen, Yingjie Wang, Dacheng TaoFirst submitted…
Evolutionary Retrofittingby Mathurin Videau, Mariia Zameshina, Alessandro Leite, Laurent Najman, Marc Schoenauer, Olivier TeytaudFirst submitted…
Beyond Linear Approximations: A Novel Pruning Approach for Attention Matrixby Yingyu Liang, Jiangxuan Long, Zhenmei…
Bypassing the Exponential Dependency: Looped Transformers Efficiently Learn In-context by Multi-step Gradient Descentby Bo Chen,…
Advancing the Understanding of Fixed Point Iterations in Deep Neural Networks: A Detailed Analytical Studyby…
QSpec: Speculative Decoding with Complementary Quantization Schemesby Juntao Zhao, Wenhao Lu, Sheng Wang, Lingpeng Kong,…