Get Started
Home
Topics
Search
Library
Research questionHow can LLMs avoid unsafe pharmacological inferences driven by drug-name affixes?Drug-name affixes can trigger class-level pharmacological responses even for fictitious drugs, leading models to produce confident but unsupported clinical content. Models may also fail to reveal when affix cues, rather than the full drug name, drive their interpretation.
AI
Alignment & Safety
Health
Mechanistic Interpretability
Natural Language Processing
Research Paper
Latest papersRecent research connected to this question, newest first.What's in a Name? Morphological Shortcuts by LLMs in PharmacologyThe evidence comes from behavioral and mechanistic analyses of 653 drugs, including fictitious names constructed from real affixes. The study distinguishes reliance on affixes, stems, or whole names and uses activation patching to localize the behavior to early-to-mid layers across models.research paper · Sep 2, 2026
Related questions
Can prompt phrasing reliably improve LLM-derived chemical features for drug-toxicity prediction?How can medical LLMs keep equivalent questions equally informative across language, register, and health-literacy cues?How can LLM agents make safe primary-care decisions as the action space grows?How can LLM-derived knowledge bases disambiguate homonymous entities and preserve auditable fact provenance?