Training atomic reasoning models on prefix and suffix tagged CoT data, then merging or multitask-combining them, improves compositional generalization on unseen skill combinations.
Werchan, Anne G.E
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
-
Learning Composable Chains-of-Thought
Training atomic reasoning models on prefix and suffix tagged CoT data, then merging or multitask-combining them, improves compositional generalization on unseen skill combinations.