Skip to content

Commit 86367dd

Browse files
committed
Add paper source and generated artifacts
- Add LaTeX source files (main.tex, sections, bibliography) - Add generated tables (12 .tex files from analysis pipeline) - Add generated figures (8 PDFs and PNGs from analysis pipeline) - Add compiled paper PDF (28 pages, 552 KB) - Add Makefile for paper compilation - Tables and figures prove programmatic reproducibility
1 parent 3f29b58 commit 86367dd

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

43 files changed

+1750
-0
lines changed

paper/Makefile

Lines changed: 13 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,13 @@
1+
MAIN = main
2+
LATEX = pdflatex
3+
4+
all: $(MAIN).pdf
5+
6+
$(MAIN).pdf: $(MAIN).tex preamble.tex macros.tex bibliography.tex
7+
$(LATEX) $(MAIN)
8+
$(LATEX) $(MAIN)
9+
10+
clean:
11+
rm -f *.aux *.bbl *.blg *.log *.out *.toc *.lof *.lot *.fls *.fdb_latexmk *.synctex.gz
12+
13+
.PHONY: all clean

paper/bibliography.tex

Lines changed: 199 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,199 @@
1+
% Shared bibliography — all venues use the same references
2+
3+
\bibliographystyle{plain}
4+
% \bibliography{references} % Uncomment when references.bib exists
5+
6+
% Temporary inline references for skeleton
7+
\begin{thebibliography}{99}
8+
9+
\bibitem{wang2023bitnet}
10+
H.~Wang et al.
11+
\newblock BitNet: Scaling 1-bit Transformers for Large Language Models.
12+
\newblock \emph{arXiv:2310.11453}, 2023.
13+
14+
\bibitem{ma2024bitnetb158}
15+
S.~Ma et al.
16+
\newblock The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits.
17+
\newblock \emph{arXiv:2402.17764}, 2024.
18+
19+
\bibitem{courbariaux2015binaryconnect}
20+
M.~Courbariaux et al.
21+
\newblock BinaryConnect: Training Deep Neural Networks with Binary Weights during Propagations.
22+
\newblock \emph{NeurIPS}, 2015.
23+
24+
\bibitem{rastegari2016xnornet}
25+
M.~Rastegari et al.
26+
\newblock XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks.
27+
\newblock \emph{ECCV}, 2016.
28+
29+
\bibitem{li2016ternary}
30+
F.~Li et al.
31+
\newblock Ternary Weight Networks.
32+
\newblock \emph{arXiv:1605.04711}, 2016.
33+
34+
\bibitem{krizhevsky2009cifar}
35+
A.~Krizhevsky.
36+
\newblock Learning Multiple Layers of Features from Tiny Images.
37+
\newblock Technical report, 2009.
38+
39+
\bibitem{devries2017cutout}
40+
T.~DeVries and G.~Taylor.
41+
\newblock Improved Regularization of Convolutional Neural Networks with Cutout.
42+
\newblock \emph{arXiv:1708.04552}, 2017.
43+
44+
\bibitem{cubuk2019autoaugment}
45+
E.~Cubuk et al.
46+
\newblock AutoAugment: Learning Augmentation Strategies from Data.
47+
\newblock \emph{CVPR}, 2019.
48+
49+
\bibitem{cubuk2020randaugment}
50+
E.~Cubuk et al.
51+
\newblock RandAugment: Practical Automated Data Augmentation with a Reduced Search Space.
52+
\newblock \emph{NeurIPS}, 2020.
53+
54+
\bibitem{he2016resnet}
55+
K.~He et al.
56+
\newblock Deep Residual Learning for Image Recognition.
57+
\newblock \emph{CVPR}, 2016.
58+
59+
\bibitem{hinton2015distilling}
60+
G.~Hinton, O.~Vinyals, and J.~Dean.
61+
\newblock Distilling the Knowledge in a Neural Network.
62+
\newblock \emph{NeurIPS Deep Learning Workshop}, 2015.
63+
64+
\bibitem{zhu2017ttq}
65+
C.~Zhu, S.~Han, H.~Mao, and W.~Dally.
66+
\newblock Trained Ternary Quantization.
67+
\newblock \emph{ICLR}, 2017.
68+
69+
\bibitem{kim2019qkd}
70+
J.~Kim, Y.~Bhalgat, J.~Lee, C.~Patel, and N.~Kwak.
71+
\newblock QKD: Quantization-aware Knowledge Distillation.
72+
\newblock \emph{arXiv:1911.12491}, 2019.
73+
74+
\bibitem{wang2019haq}
75+
K.~Wang, Z.~Liu, Y.~Lin, J.~Lin, and S.~Han.
76+
\newblock HAQ: Hardware-Aware Automated Quantization with Mixed Precision.
77+
\newblock \emph{CVPR}, 2019.
78+
79+
\bibitem{nielsen2024bitnetreloaded}
80+
J.~Nielsen and P.~Schneider-Kamp.
81+
\newblock BitNet b1.58 Reloaded: State-of-the-art Performance Also on Smaller Networks.
82+
\newblock \emph{arXiv:2407.09527}, 2024.
83+
84+
\bibitem{kim2025bdnet}
85+
D.~Kim, J.-S.~Lee, N.-r.~Kim, S.~Lee, and J.-H.~Lee.
86+
\newblock BD-Net: Has Depth-Wise Convolution Ever Been Applied in Binary Neural Networks?
87+
\newblock \emph{AAAI}, 2026.
88+
89+
\bibitem{zhou2016dorefa}
90+
S.~Zhou, Y.~Wu, Z.~Ni, X.~Zhou, H.~Wen, and Y.~Zou.
91+
\newblock DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients.
92+
\newblock \emph{arXiv:1606.06160}, 2016.
93+
94+
\bibitem{dong2019hawq}
95+
Z.~Dong, Z.~Yao, A.~Gholami, M.~Mahoney, and K.~Keutzer.
96+
\newblock HAWQ: Hessian AWare Quantization of Neural Networks with Mixed-Precision.
97+
\newblock \emph{ICCV}, 2019.
98+
99+
\bibitem{elthakeb2020dcq}
100+
A.~T.~Elthakeb, P.~Pilligundla, F.~Mireshghallah, A.~Alaghi, and H.~Esmaeilzadeh.
101+
\newblock Divide and Conquer: Leveraging Intermediate Feature Representations for Quantized Training of Neural Networks.
102+
\newblock \emph{arXiv:1906.06033}, 2020.
103+
104+
\bibitem{le2015tinyimagenet}
105+
Y.~Le and X.~Yang.
106+
\newblock Tiny ImageNet Visual Recognition Challenge.
107+
\newblock CS 231N, Stanford University, 2015.
108+
109+
\bibitem{liu2020reactnet}
110+
Z.~Liu, Z.~Shen, M.~Savvides, and K.-T.~Cheng.
111+
\newblock ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions.
112+
\newblock \emph{ECCV}, 2020.
113+
114+
\bibitem{guo2022bnext}
115+
N.~Guo, J.~Bethge, C.~Meinel, and H.~Yang.
116+
\newblock BNext: Any Precision Binary Neural Network.
117+
\newblock \emph{arXiv:2211.12933}, 2022.
118+
119+
\bibitem{dong2020hawqv2}
120+
Z.~Dong, Z.~Yao, D.~Arfeen, A.~Gholami, M.~W.~Mahoney, and K.~Keutzer.
121+
\newblock HAWQ-V2: Hessian Aware trace-Weighted Quantization of Neural Networks.
122+
\newblock \emph{NeurIPS}, 2020.
123+
124+
\bibitem{qin2020irnet}
125+
H.~Qin, R.~Gong, X.~Liu, M.~Shen, Z.~Wei, F.~Yu, and J.~Song.
126+
\newblock Forward and Backward Information Retention for Accurate Binary Neural Networks.
127+
\newblock \emph{CVPR}, 2020.
128+
129+
\bibitem{zhao2024sqakd}
130+
K.~Zhao and M.~Zhao.
131+
\newblock SQAKD: Self-supervised Quantization-Aware Knowledge Distillation.
132+
\newblock \emph{AISTATS}, 2024.
133+
134+
\bibitem{bengio2013estimating}
135+
Y.~Bengio, N.~Léonard, and A.~Courville.
136+
\newblock Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation.
137+
\newblock \emph{arXiv:1308.3432}, 2013.
138+
139+
\bibitem{kim2021relaxloss}
140+
J.~Kim, S.~Bhattacharjee, S.~Park, S.~Jung, and Y.~M.~Kim.
141+
\newblock RelaxLoss: Penalty-Free Quantization through Normalized Loss.
142+
\newblock \emph{arXiv:2105.00944}, 2021.
143+
144+
\bibitem{zhang2018mixup}
145+
H.~Zhang, M.~Cisse, Y.~N.~Dauphin, and D.~Lopez-Paz.
146+
\newblock mixup: Beyond Empirical Risk Minimization.
147+
\newblock \emph{ICLR}, 2018.
148+
149+
\bibitem{gundersen2018reproducibility}
150+
O.~E.~Gundersen and S.~Kjensmo.
151+
\newblock State of the Art: Reproducibility in Artificial Intelligence.
152+
\newblock \emph{AAAI}, 2018.
153+
154+
\bibitem{hutson2018ai}
155+
M.~Hutson.
156+
\newblock Artificial Intelligence Faces Reproducibility Crisis.
157+
\newblock \emph{Science}, 359(6377):725--726, 2018.
158+
159+
\bibitem{dodge2019mlchecklist}
160+
J.~Dodge, S.~Gururangan, D.~Card, R.~Schwartz, and N.~A.~Smith.
161+
\newblock Show Your Work: Improved Reporting of Experimental Results.
162+
\newblock \emph{EMNLP}, 2019.
163+
164+
\bibitem{tflite}
165+
TensorFlow Lite Team.
166+
\newblock TensorFlow Lite: Deploy machine learning models on mobile and edge devices.
167+
\newblock \url{https://www.tensorflow.org/lite}, 2023.
168+
169+
\bibitem{pytorch_mobile}
170+
PyTorch Team.
171+
\newblock PyTorch Mobile: End-to-end workflow from training to deployment.
172+
\newblock \url{https://pytorch.org/mobile/}, 2023.
173+
174+
\bibitem{bitblas}
175+
L.~Wang, L.~Lei, L.~Ye, Y.~Zhao, W.~Chen, D.~Lin, X.~Zheng, and C.~Yu.
176+
\newblock BitBLAS: A High-Performance Library for Quantized Deep Learning.
177+
\newblock \emph{arXiv:2410.16144}, 2024.
178+
179+
\bibitem{pytorch_cifar_kuangliu}
180+
K.~Liu.
181+
\newblock Train CIFAR10 with PyTorch.
182+
\newblock \url{https://github.com/kuangliu/pytorch-cifar}, 2017.
183+
184+
\bibitem{pytorch_cifar100_weiaicunzai}
185+
W.~Zhang.
186+
\newblock PyTorch CIFAR-100 Benchmark.
187+
\newblock \url{https://github.com/weiaicunzai/pytorch-cifar100}, 2019.
188+
189+
\bibitem{tiny_imagenet_benchmark}
190+
Y.~Le and X.~Yang.
191+
\newblock Tiny ImageNet Visual Recognition Challenge.
192+
\newblock CS 231N, Stanford University, 2015.
193+
194+
\bibitem{cover2006information}
195+
T.~M.~Cover and J.~A.~Thomas.
196+
\newblock Elements of Information Theory, 2nd Edition.
197+
\newblock Wiley-Interscience, 2006.
198+
199+
\end{thebibliography}

paper/figures/.gitkeep

Whitespace-only changes.
16.5 KB
Binary file not shown.
33.7 KB
Loading
18.1 KB
Binary file not shown.
37.8 KB
Loading
31.4 KB
Binary file not shown.
146 KB
Loading

paper/figures/gap_scaling.pdf

27.6 KB
Binary file not shown.

0 commit comments

Comments
 (0)