Fetching the paper…
Reading the bibliography…
Traditional Deep Neural Network (DNN) quantization methods using integer, fixed-point, or floating-point data types struggle to capture diverse DNN parameter distributions at low precision, and often require large silicon overhead and intensive quantization-aware training.
On the meaning and use of kurtosis
Lawrence T DeCarlo. 1997 · 1997
Earlier work this paper cites.
From high-level deep neural models to FPGAs. In
H. Sharma et al · 2016
Earlier work this paper cites.
Beating floating point at its own game: Posit arithmetic
J. Gustafson and I. Yonemoto. 2017 · 2017
Earlier work this paper cites.
Bit fusion: Bit-level dynamically composable architecture for accelerating deep neural network. In
H. Sharma et al · 2018
Earlier work this paper cites.
Cheetah: Mixed low-precision hardware & software co-design framework for DNNs on the edge
H. Langroudi et al · 2019
Earlier work this paper cites.
Positnn framework: Tapered precision deep learning inference for the edge. In
H.F. Langroudi et al · 2019
Earlier work this paper cites.
Zeroq A novel zero shot quantization framework. In
Y. Cai et al · 2020
Earlier work this paper cites.
Deep PeNSieve: A deep learning framework based on the posit number system
R. Murillo et al · 2020
Cited alongside, same era.
Algorithm-hardware co-design of adaptive floating-point encodings for resilient deep learning inference. In
T. Tambe et al · 2020
Cited alongside, same era.
Low-precision logarithmic number systems: beyond base-2
S.A. Alam et al · 2021
Cited alongside, same era.
Brecq: Pushing the limit of post-training quantization by block reconstruction
Y. Li et al · 2021
Cited alongside, same era.
Fq-vit: Post-training quantization for fully quantized vision transformer
Y. Lin et al · 2021
Cited alongside, same era.
Hawq-v3: Dyadic neural network quantization. In
Z. Yao et al · 2021
Later among the works it cites.
Robustness to Adversarial Gradients: A Glimpse Into the Loss Landscape of Contrastive Pre-training. In
P. Fradkin et al · 2022
Later among the works it cites.
Ant: Exploiting adaptive numerical data type for low-bit deep neural network quantization. In
C. Guo et al · 2022
Later among the works it cites.
PositIV: A Configurable Posit Processor Architecture for Image and Video Processing. In
A. Ramachandran et al · 2022
Later among the works it cites.
EMQ: Evolving Training-free Proxies for Automated Mixed Precision Quantization. In
P. Dong et al · 2023
Later among the works it cites.
Jumping through Local Minima: Quantization in the Loss Landscape of Vision Transformers. In
alphaXiv searches the wider corpus for related work and actual follow-ups.
alphaXiv is searching for related work…
Improving neural network efficiency via post-training quantization with adaptive floating-point. In
F. Liu et al · 2021
Cited alongside, same era.
DeepScaleTool: A tool for the accurate estimation of technology scaling in the deep-submicron era. In
S. Sarangi et al · 2021
Cited alongside, same era.
N. Frumkin et al · 2023
Later among the works it cites.
A 95.6-TOPS/W Deep Learning Inference Accelerator With Per-Vector Scaled 4-bit Quantization in 5 nm
B. Keller et al · 2023
Later among the works it cites.