Contextual Earnings-22 is a new benchmark dataset showing that scaled keyword prompting and boosting both deliver significantly better accuracy on custom vocabularies than standard academic tests.
Binary codes capable of correcting deletions, insertions, and reversals,
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
-
Contextual Earnings-22: A Speech Recognition Benchmark with Custom Vocabulary in the Wild
Contextual Earnings-22 is a new benchmark dataset showing that scaled keyword prompting and boosting both deliver significantly better accuracy on custom vocabularies than standard academic tests.