/AI7h ago

Critics Question Temperature Zero Practice In LLM Benchmark Papers

--0--
Quote posts
Comments
Original post
Aryaman Arora@aryaman2020#678inAI

i never wanted to do LLM API research because it all feels like cargo-culting. no one can explain why x procedure is the right thing to do but still clings to passionate opinions (e.g. this thread and replies)

3:29 PM · Jun 1, 2026 · 11.2K Views
Sentiment
Sentiment unavailable for this story.
Cluster Engagement
-
Views
-
Comments
-
Reposts
-
Bookmarks
Expand data
Posts from X
Most Activity
Most ActivityTimeline
VIEWS953LIKES7
Aryaman Arora@aryaman2020

maybe you can only really do such work rigourously inside a lab, where the internal settings are at least transparent

Aryaman Arora@aryaman2020

i never wanted to do LLM API research because it all feels like cargo-culting. no one can explain why x procedure is the right thing to do but still clings to passionate opinions (e.g. this thread and replies)

7hViews 953Likes 7Bookmarks 0