CAP-CoT improves LLM reasoning accuracy and stability by iteratively refining solver prompts via contrast with adversarially generated flawed reasoning chains.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.soc-ph 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
No citing papers match the current filters.