Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
boodleboodle
40 days ago
|
parent
|
context
|
favorite
| on:
Tracing the thoughts of a large language model
This is why, whenever I can, I call RLHF/DPO "sequence level calibration" instead of "alignment tuning".
Some precursors to RLHF:
https://arxiv.org/abs/2210.00045
https://arxiv.org/abs/2203.16804
Consider applying for YC's Summer 2025 batch! Applications are open till May 13
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
Some precursors to RLHF: https://arxiv.org/abs/2210.00045 https://arxiv.org/abs/2203.16804