Self-contradictory Hallucinations of Large Language Models: Evaluation, Detection and Mitigation

Our summary

Defines, detects, and measures self-contradiction — two mutually inconsistent statements produced within the same context — across LLMs, and offers detection and mitigation without external knowledge.

Why it matters

Internal consistency is a checkable reliability property; frequent self-contradiction undercuts trust even when no external fact is at stake.

Related findings (1)

• Self-contradiction within a single conversationLow

  Models assert one fact and later assert its opposite within the same session.

Cite this