Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
anotherhue
11 months ago
|
parent
|
context
|
favorite
| on:
TL;DR of Deep Dive into LLMs Like ChatGPT by Andre...
The RL-only (no SFT) approaches might remove that issue. Problem sets should be smaller (and mechanically creatable) than the entire western corpus.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
