6 papers
Variational Diffusion Channel Decoder
Chengwei Zhang, Yifan Du, Siyu Liao
Neural channel decoder, as a data-driven channel decoding strategy, has shown very promising improvement on error-correcting capability over the classical methods. However, the suc…
Beyond Geometry: Efficient Topologically-Grounded Navigation in Complex 3D Environments
Yifan Du, Chengwei Zhang, Siyu Liao +1
Ground robot navigation in complex 3D environments is often hindered by geometric ambiguity, where non-traversable structures such as furniture share local geometric properties wit…
Memory-Efficient Training with In-Place FFT Implementation
Xinyu Ding, Bangtian Liu, Siyu Liao +1
Fast Fourier Transforms (FFT) are widely used to reduce memory and computational costs in deep learning. However, existing implementations, including standard FFT and real FFT (rFF…
CDFlow: Building Invertible Layers with Circulant and Diagonal Matrices
Xuchen Feng, Siyu Liao
Normalizing flows are deep generative models that enable efficient likelihood estimation and sampling through invertible transformations. A key challenge is to design linear layers…
Block Circulant Adapter for Large Language Models
Xinyu Ding, Meiqi Wang, Siyu Liao +1
Fine-tuning large language models (LLMs) is difficult due to their huge model size. Recent Fourier domain-based methods show potential for reducing fine-tuning costs. We propose a…
Parameter-Efficient Fine-Tuning with Circulant and Diagonal Vectors
Xinyu Ding, Lexuan Chen, Siyu Liao +1
Foundation models have achieved tremendous success in different domains. However, their huge computation and storage complexity make these models difficult to fine-tune and also le…