Y
Hacker News
new
|
ask
|
show
|
jobs
by
anonzzzies
163 days ago
We detect any use of any and the LLM has to fix them before our check succeeds. It does and works fine.
1 comments
akshay326
163 days ago
currently starting to do the same over seer's frontend, i didn't realise how simple yet effective this technique / guardrail could be!
link