Parallel Algorithms for Exact Enumeration of Deep Neural Network Activation Regions
A feedforward neural network using rectified linear units constructs a mapping from inputs to outputs by partitioning its input space into a set of convex regions where points within a region share a single affine transformation. In order to understand how neural networks work, when and why they fail, and how they compare to biological intelligence, we need to understand the organization and formation of these regions. Step one is to design and implement algorithms for exact region enumeration in networks beyond toy examples. In this work, we present parallel algorithms for exact enumeration in deep (and shallow) neural networks. Our work has three main contributions: (1) we present a novel algorithm framework and parallel algorithms for region enumeration; (2) we implement one of our algorithms on a variety of network architectures and experimentally show how the number of regions dictates runtime; and (3) we show, using our algorithm's output, how the dimension of a region's affine transformation impacts further partitioning of the region by deeper layers. To our knowledge, we run our implemented algorithm on networks larger than all of the networks used in the existing region enumeration literature. Further, we experimentally demonstrate the importance of parallelism for region enumeration of any reasonably sized network..
| Medienart: |
|
|---|
Preprint| Erscheinungsjahr: |
2024 |
|---|---|
| Erschienen: |
2024 |
| Enthalten in: |
arXiv.org - (2024) vom: 29. Feb. Zur Gesamtaufnahme - year:2024 |
|---|
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Drammis, Sabrina [VerfasserIn] |
|---|
| Links: |
Volltext [kostenfrei] |
|---|
| Themen: |
000 |
|---|
| Förderinstitution / Projekttitel: |
|
|---|
| PPN (Katalog-ID): |
XCH042761557 |
|---|
| LEADER | 01000naa a22002652 4500 | ||
|---|---|---|---|
| 001 | XCH042761557 | ||
| 003 | DE-627 | ||
| 005 | 20240306114445.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 240306s2024 xx |||||o 00| ||eng c | ||
| 035 | |a (DE-627)XCH042761557 | ||
| 035 | |a (chemrXiv)2403.00860 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Drammis, Sabrina |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Parallel Algorithms for Exact Enumeration of Deep Neural Network Activation Regions |
| 264 | 1 | |c 2024 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 520 | |a A feedforward neural network using rectified linear units constructs a mapping from inputs to outputs by partitioning its input space into a set of convex regions where points within a region share a single affine transformation. In order to understand how neural networks work, when and why they fail, and how they compare to biological intelligence, we need to understand the organization and formation of these regions. Step one is to design and implement algorithms for exact region enumeration in networks beyond toy examples. In this work, we present parallel algorithms for exact enumeration in deep (and shallow) neural networks. Our work has three main contributions: (1) we present a novel algorithm framework and parallel algorithms for region enumeration; (2) we implement one of our algorithms on a variety of network architectures and experimentally show how the number of regions dictates runtime; and (3) we show, using our algorithm's output, how the dimension of a region's affine transformation impacts further partitioning of the region by deeper layers. To our knowledge, we run our implemented algorithm on networks larger than all of the networks used in the existing region enumeration literature. Further, we experimentally demonstrate the importance of parallelism for region enumeration of any reasonably sized network. | ||
| 650 | 4 | |a Computer Science - Machine Learning |7 (dpeaa)DE-84 | |
| 650 | 4 | |a Computer Science - Artificial Intelligence |7 (dpeaa)DE-84 | |
| 650 | 4 | |a Computer Science - Neural and Evolutionary Computing |7 (dpeaa)DE-84 | |
| 650 | 4 | |a 000 |7 (dpeaa)DE-84 | |
| 700 | 1 | |a Zheng, Bowen |4 aut | |
| 700 | 1 | |a Srinivasan, Karthik |4 aut | |
| 700 | 1 | |a Berwick, Robert C. |4 aut | |
| 700 | 1 | |a Lynch, Nancy A. |4 aut | |
| 700 | 1 | |a Ajemian, Robert |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t arXiv.org |g (2024) vom: 29. Feb. |
| 773 | 1 | 8 | |g year:2024 |g day:29 |g month:02 |
| 856 | 4 | 0 | |u https://arxiv.org/abs/2403.00860 |z kostenfrei |3 Volltext |
| 912 | |a GBV_XCH | ||
| 951 | |a AR | ||
| 952 | |j 2024 |b 29 |c 02 | ||