DeepSeek-R1's final outputs omit sensitive topic keywords that appear in its internal chain-of-thought, indicating semantic-level information suppression.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CY 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Information Suppression in Large Language Models: Auditing, Quantifying, and Characterizing Censorship in DeepSeek
DeepSeek-R1's final outputs omit sensitive topic keywords that appear in its internal chain-of-thought, indicating semantic-level information suppression.