AIBullisharXiv – CS AI · 7h ago6/10
🧠
LLMs Need Encoders for Semantic IDs Too
Researchers propose PrefixMem, a dedicated encoder for Semantic IDs (hierarchical codes used in generative recommendation systems), arguing that LLMs require specialized preprocessing for this modality just as they do for vision and audio. Testing at Pinterest shows accuracy improvements up to 46% and retrieval recall gains of 22%, particularly on difficult cases where standard decoding fails.