Empowering Compact LLMs with Fusion of Layer-wise Exits for Recommendation

2026-08-17

AI Summary

为解决紧凑型大语言模型推荐系统表达能力和适应性不足的问题，提出FLEXRec方法，通过在多层插入预测头并自适应融合其分数分布来增强模型性能。

Abstract

Large language model-based recommender systems (LLM-RSs) have demonstrated remarkable capabilities, but are computationally unsustainable for many real-world applications. Compact LLMs offer a practical alternative, yet their reduced capacity often requires reasoning or knowledge distillation methods that increase latency or depend on larger models. Combined with autoregressive generation, these approaches face severe scalability bottlenecks. In contrast, discriminative LLM-RSs enable efficient full-corpus ranking through embedding similarity, but compact backbones remain limited in expressiveness and structural adaptivity. We propose the Fusion of Layer-wise Exits for Sequential Recommendation (FLEXRec), a discriminative framework that enhances compact LLMs while retaining scalable full-corpus ranking. FLEXRec inserts prediction heads (i.e., exits) at multiple transformer layers and adaptively fuses their score distributions. An adaptive continuous router (AC-Router) dynamically selects both the number and identity of exits for each user sequence, while a novel target-k hinge loss regulates routing sparsity. Experiments on three real-world datasets with Qwen 3 1.7B and Llama 3.2 3B show that FLEXRec achieves state-of-the-art accuracy among compact-backbone methods while remaining highly efficient. Code: https://github.com/xurong-liang/FLEXRec

Problem

• Large Language Models
• Recommender Systems
• Compact LLMs
• Scalability
• Efficiency

Innovation

• FLEXRec
• Layer-wise Exits
• Adaptive Continuous Router (AC-Router)
• Target-k Hinge Loss