Transformer models trained from scratch on in-context linear function prediction learn boundary-limited interpolation, not general linear regression.
Advances in Neural Information Processing Systems 36, 45614–45650 (2023)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Re-examining learning linear functions in context
Transformer models trained from scratch on in-context linear function prediction learn boundary-limited interpolation, not general linear regression.