AI evals are becoming the new compute bottleneck
•
16
AI & ML interests
We’re building a research coalition on evaluating evaluations (EvalEval)! Hosted by Hugging Face, University of Edinburgh, and EleutherAI.
Recent Activity
View all activity
Articles
evaleval/EEE_datastore
Updated • 3.57k • 23
evaleval/card_backend
Updated • 13.1k
evaleval/entity-registry-data
Viewer • Updated • 11 • 111
evaleval/auto-benchmarkcards
Preview • Updated • 413 • 3
evaleval/alphaxiv_datastore
Updated • 19 • 1
evaleval/social_impact_eval_annotations
Viewer • Updated • 4.24k • 48 • 4
evaleval/old_eee
Preview • Updated • 7