Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
onceonceonce
23 days ago
|
parent
|
context
|
favorite
| on:
Disagreement among frontier LLMs on real-world fac...
Teams I work with use the abstain rate to flag what goes to a human. Disagreement between models is the same idea. Your 67% is what makes "two cheap models, escalate when they fight" actually work. Without abstain it mostly looks like noise.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: