Academic Search Engines (ASE) help researchers discover literature but mostly rely on keyword searches and static filters such as date or citation count, which operate only on metadata. Platforms like the Open Research Knowledge Graph (ORKG) and ORKG-Ask, though advanced, face similar limitations. To overcome this, we introduce Smart Filters, a context-aware feature integrated into both systems. Smart Filters use a neuro-symbolic approach that combines Knowledge Graphs and Large Language Models to dynamically generate semantic facets from paper content. Our evaluation shows that Smart Filters improve efficiency, relevance, and the overall user search experience.