After removing label errors from fifteen standard benchmarks, frontier LLMs still fail simple tasks, so current benchmarks measure capability but not reliability.
Parsing algebraic word problems into equations
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Do Large Language Model Benchmarks Test Reliability?
After removing label errors from fifteen standard benchmarks, frontier LLMs still fail simple tasks, so current benchmarks measure capability but not reliability.