Chain-of-thought tokens in a fine-tuned 1.5B language model behave like program variables: they store intermediate values, can be replaced by latent vectors without loss, and intervening on their values causally changes the output.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Chain-of-Thought Tokens are Computer Program Variables
Chain-of-thought tokens in a fine-tuned 1.5B language model behave like program variables: they store intermediate values, can be replaced by latent vectors without loss, and intervening on their values causally changes the output.