{"ID":2854166,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.15729","arxiv_id":"2510.15729","title":"FACE: A General Framework for Mapping Collaborative Filtering Embeddings into LLM Tokens","abstract":"Recently, large language models (LLMs) have been explored for integration with collaborative filtering (CF)-based recommendation systems, which are crucial for personalizing user experiences. However, a key challenge is that LLMs struggle to interpret the latent, non-semantic embeddings produced by CF approaches, limiting recommendation effectiveness and further applications. To address this, we propose FACE, a general interpretable framework that maps CF embeddings into pre-trained LLM tokens. Specifically, we introduce a disentangled projection module to decompose CF embeddings into concept-specific vectors, followed by a quantized autoencoder to convert continuous embeddings into LLM tokens (descriptors). Then, we design a contrastive alignment objective to ensure that the tokens align with corresponding textual signals. Hence, the model-agnostic FACE framework achieves semantic alignment without fine-tuning LLMs and enhances recommendation performance by leveraging their pre-trained capabilities. Empirical results on three real-world recommendation datasets demonstrate performance improvements in benchmark models, with interpretability studies confirming the interpretability of the descriptors. Code is available in https://github.com/YixinRoll/FACE.","short_abstract":"Recently, large language models (LLMs) have been explored for integration with collaborative filtering (CF)-based recommendation systems, which are crucial for personalizing user experiences. However, a key challenge is that LLMs struggle to interpret the latent, non-semantic embeddings produced by CF approaches, limit...","url_abs":"https://arxiv.org/abs/2510.15729","url_pdf":"https://arxiv.org/pdf/2510.15729v1","authors":"[\"Chao Wang\",\"Yixin Song\",\"Jinhui Ye\",\"Chuan Qin\",\"Dazhong Shen\",\"Lingfeng Liu\",\"Xiang Wang\",\"Yanyong Zhang\"]","published":"2025-10-17T15:19:54Z","proceeding":"cs.IR","tasks":"[\"cs.IR\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false,"code_links":[{"ID":608122,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2854166,"paper_url":"https://arxiv.org/abs/2510.15729","paper_title":"FACE: A General Framework for Mapping Collaborative Filtering Embeddings into LLM Tokens","repo_url":"https://github.com/YixinRoll/FACE","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}