An empirical audit shows the translation tax in English-to-Chinese benchmarks is estimator- and item-dependent rather than a single scalar effect, with a residue dose-response in naturalization tests.
Measuring massive multitask language understanding
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CL 2representative citing papers
Massive activations are constant large values in LLMs that function as indispensable bias terms and concentrate attention probabilities on specific tokens.
citing papers explorer
-
The Translation Tax Is Not a Scalar: A Counterfactual Audit of English-Source Cue Inheritance in Chinese Multilingual Benchmarks
An empirical audit shows the translation tax in English-to-Chinese benchmarks is estimator- and item-dependent rather than a single scalar effect, with a residue dose-response in naturalization tests.
-
Massive Activations in Large Language Models
Massive activations are constant large values in LLMs that function as indispensable bias terms and concentrate attention probabilities on specific tokens.