Timeline for Why does DNABert use overlapping k-mers as input?
Current License: CC BY-SA 4.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 19, 2023 at 12:17 | history | edited | M__♦ | CC BY-SA 4.0 |
deleted 25 characters in body
|
| Sep 19, 2023 at 0:15 | history | edited | M__♦ | CC BY-SA 4.0 |
added 53 characters in body
|
| Sep 19, 2023 at 0:01 | comment | added | M__♦ | I've tried my best above. | |
| Sep 19, 2023 at 0:01 | history | edited | M__♦ | CC BY-SA 4.0 |
added 543 characters in body
|
| Sep 18, 2023 at 23:49 | history | edited | M__♦ | CC BY-SA 4.0 |
added 543 characters in body
|
| Sep 18, 2023 at 19:13 | history | edited | M__♦ | CC BY-SA 4.0 |
added 185 characters in body
|
| Sep 16, 2023 at 21:51 | comment | added | Avatrin | I am sorry, but this does not answer my question regarding why one would use k-mers for k>1 versus 1-mers with positional encoding as is the norm with transformers. Not just BERT and other NLP transformer-like models (aka LLMs) but even ViT uses positional encoding | |
| Sep 16, 2023 at 21:48 | history | edited | M__♦ | CC BY-SA 4.0 |
added 17 characters in body
|
| Sep 16, 2023 at 21:27 | history | answered | M__♦ | CC BY-SA 4.0 |