Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
Noumenon72
61 days ago
|
parent
|
context
|
favorite
| on:
Defeating Nondeterminism in LLM Inference
Are the results of the matmuls really that far apart in size that you have to lose significant bits when adding them up at FP32?
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
