arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2611.10285cs.AI
收藏

面向大语言模型推荐器的保形公平解码

FAIDER: Fairness-Aware Item-Level Decoding for Large Language Model Recommenders

University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)

Ziwei Wu, Jingrui He

AI总结:

本文针对大语言模型推荐器在权重与全词汇对数几率不可用时的公平性控制问题,提出Fair-Rec解码框架,经实验验证其可降低反事实差异并保持推荐性能。

摘要:

大语言模型(LLM)无需针对特定任务训练即可生成个性化推荐,但在保持交互历史固定的情况下改变受保护属性可能会改变推荐项的语义内容。本文研究了在模型权重和全词汇对数几率不可用时的公平性控制问题,提出了Fair-Rec,这是一种项级解码框架,它使用离线校准的语义差异阈值,并在生成推荐列表时控制候选项的选择。硬变体根据阈值过滤候选项;软变体权衡候选项的似然性和差异,同时利用观察到的违规情况调整惩罚。在显式统一可执行性假设下,我们为软解码器的经验违规率超额建立了路径式O(1/T)上界。在MovieLens-1M和Amazon Movies & TV数据集上使用Qwen3-8B和Llama3.1-8B-Instruct进行的实验表明,与零样本推荐、UP5和FACTER相比,软变体降低了测量的反事实差异,同时保持了具有竞争力的HR@10和NDCG@10。

英文摘要:

Large language model (LLM) recommenders can produce semantically different recommendations when a user's protected attribute changes, even if the interaction history remains fixed. Controlling this disparity is challenging when model weights and full-vocabulary logits are inaccessible. We propose FAIDER (Fairness-Aware Item-level DEcoding for Recommendation), which incorporates fairness into list construction by selecting complete item candidates rather than manipulating token-level logits or repeatedly repairing finished outputs. The central idea is to separate the operational criterion for a violation from how strongly generation is constrained. Conformal-inspired empirical calibration fixes a cross-group semantic-disparity threshold, while online decoding applies either hard filtering or a soft penalty adjusted through dual-inspired feedback to target a prescribed violation rate. Separately, we establish a pathwise guarantee: under a uniform enforceability assumption, the soft decoder's empirical violation rate exceeds its target by at most $O(1/T)$. Experiments across recommendation benchmarks and LLM backbones show that Soft FAIDER improves both measured semantic fairness and ranking quality over the evaluated prompt-repair baselines. It also generally achieves a more favorable fairness-utility balance than the hard variant, indicating that stricter per-recommendation constraints do not necessarily yield lower measured disparity.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑