Timeline for Why use ReLU over Leaky ReLU?
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 14, 2023 at 19:31 | comment | added | John Brown | @JobHunter69 A sparse activation function is more strict when it comes to letting non-zero activations through than a non-sparse function. This is beneficial b/c it leads to a more generalizable model that "focuses" on only the most distinguishing features of the data. Read Deep Sparse Rectifier Neural Networks by Bengio et al. for a more thorough explanation of the topic. | |
| Nov 12, 2023 at 17:53 | comment | added | JobHunter69 | Can I ask what you mean by sparsity? | |
| May 26, 2023 at 14:41 | vote | accept | John Brown | ||
| May 25, 2023 at 10:13 | answer | added | Luca Anzalone | timeline score: 5 | |
| S May 24, 2023 at 21:44 | review | First questions | |||
| May 25, 2023 at 10:15 | |||||
| S May 24, 2023 at 21:44 | history | asked | John Brown | CC BY-SA 4.0 |