Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
CamperBob2
35 days ago
|
parent
|
context
|
favorite
| on:
How attention sinks keep language models stable
Good point -- I can see that, but it all ends up in the same context, anyway. Point being, the model seems to prefer to conserve tokens.
That said, now I'm wondering if all those dashes it spews out are more than just window dressing.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
That said, now I'm wondering if all those dashes it spews out are more than just window dressing.