A study investigates how different forms of conceptual grounding affect antisemitism detection and explanation behavior across four state-of-the-art large language models. Using two expert-annotated datasets, the researchers compared definitional, fine-grained taxonomic, example-augmented, and large-context representations of antisemitism.
- Fine-grained taxonomic representations substantially improve recall while simultaneously reducing precision.
- Supplying substantially larger conceptual resources yields no additional quantitative benefit.
- Post-Holocaust antisemitism poses the most persistent challenge across models and configurations.
- Analysis reveals systematic limitations including overproduction of conceptual references, reliance on lexical cues, overconfidence, and difficulties with subtle or justificatory forms of antisemitism.
The findings highlight both the potential and the remaining limitations of conceptually grounded LLMs for detecting and reasoning about antisemitism.