These LLMs are the best at resisting Russian propaganda

The Estonian Language Institute (ELI) has developed a 'Propaganda Resistance' benchmark to evaluate large language models' (LLMs) ability to avoid Russian Federation narratives on topics like Crimea, Ukraine, and NATO. The benchmark, created with Propastop, assesses LLM responses to neutral, biased, or malicious questions in multiple languages, using an AI model calibrated to expert judgments.