Empirical Jacobian analysis reveals that token influence in trained language models decays as a power law with distance (exponent ~0.8), a learned property not present in random models.
Machine learning and domain decomposition methods - a survey.Computational Science and Engineering, 1(2):2, 2024
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
How Token Influence Decays with Distance: A Green-Function View of Trained Language Models
Empirical Jacobian analysis reveals that token influence in trained language models decays as a power law with distance (exponent ~0.8), a learned property not present in random models.