Learning Federated Representations And Recommendations With Limited Negatives

Abstract

Deep retrieval models are widely used for learning entity representations and recommendations. Federated learning provides a privacy-preserving way to train these models without requiring centralization of user data. However, federated deep retrieval models usually perform much worse than their centralized counterparts due to non-IID (independent and identically distributed) training data on clients, an intrinsic property of federated learning that limits negatives available for training. We demonstrate that this issue is distinct from the commonly studied client drift problem. This work proposes batch-insensitive losses as a way to alleviate the non-IID negatives issue for federated movie recommendations. We explore a variety of techniques and identify that batch-insensitive losses can effectively improve the performance of federated deep retrieval models, increasing the relative recall of the federated model by up to 93.15% and reducing the relative gap in recall between it and a cen

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyning2021learning

Related papers

• Lightfr: Lightweight Federated Recommendation With Privacy-preserving Matrix Factorization (2022)12.87
• Plug-and-play Parameter-efficient Tuning Of Embeddings For Federated Recommendation (2025)0.95
• Learning Effective Representations For Retrieval Using Self-distillation With Adaptive Relevance Margins (2024)2.26
• Deep Retrieval: Learning A Retrievable Structure For Large-scale Recommendations (2020)0.00
• Optimizing Dense Retrieval Model Training With Hard Negatives (2021)16.34
• Teaching Dense Retrieval Models To Specialize With Listwise Distillation And LLM Data Augmentation (2025)0.00
• L^2R: Lifelong Learning For First-stage Retrieval With Backward-compatible Representations (2023)5.24
• Hard Negatives, Hard Lessons: Revisiting Training Data Quality For Robust Information Retrieval With Llms (2025)2.26