LLM embeddings enhance auto-bidding strategies by integrating semantic and numeric data

A new research paper explores the integration of Large Language Models (LLMs) into auto-bidding systems for real-time advertising. The study found that while LLM embeddings offer valuable semantic information, they cannot entirely replace numerical features and require careful integration rather than simple concatenation. The proposed SemBid framework injects LLM-encoded semantics as tokens alongside numerical data, improving controllability and generalization across various objectives and outperforming existing baselines. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT This research could lead to more sophisticated and controllable automated bidding strategies in advertising by leveraging LLM capabilities.

RANK_REASON This is a research paper published on arXiv detailing a new framework for auto-bidding. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• LLM
• SemBid
• arXiv

• paper
• product
• other