BAGEL is a new closed-book benchmark that tests language models on diverse aspects of animal knowledge using curated and generated questions from scientific databases.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
BAGEL: Benchmarking Animal Knowledge Expertise in Language Models
BAGEL is a new closed-book benchmark that tests language models on diverse aspects of animal knowledge using curated and generated questions from scientific databases.