AsCAN: Asymmetric Convolution-Attention Networks for Efficient Recognition and Generationby Anil Kag, Huseyin Coskun, Jierun Chen,…
WiFlexFormer: Efficient WiFi-Based Person-Centric Sensingby Julian Strohmayer, Matthias Wödlinger, Martin KampelFirst submitted to arxiv on:…
Interactions Across Blocks in Post-Training Quantization of Large Language Modelsby Khasmamad Shabanovi, Lukas Wiest, Vladimir…
How Transformers Solve Propositional Logic Problems: A Mechanistic Analysisby Guan Zhe Hong, Nishanth Dikkala, Enming…
LASER: Attention with Exponential Transformationby Sai Surya Duvvuri, Inderjit S. DhillonFirst submitted to arxiv on:…
Rethinking Decoders for Transformer-based Semantic Segmentation: A Compression Perspectiveby Qishuai Wen, Chun-Guang LiFirst submitted to…
Can Transformers Smell Like Humans?by Farzaneh Taleb, Miguel Vasco, Antônio H. Ribeiro, Mårten Björkman, Danica…
Pre-trained Visual Dynamics Representations for Efficient Policy Learningby Hao Luo, Bohan Zhou, Zongqing LuFirst submitted…
Enhancing Transformer Training Efficiency with Dynamic Dropoutby Hanrui Yan, Dan ShaoFirst submitted to arxiv on:…
Pretrained transformer efficiently learns low-dimensional target functions in-contextby Kazusato Oko, Yujin Song, Taiji Suzuki, Denny…