# On component interactions in two-stage recommender systems

arXiv: https://arxiv.org/abs/2106.14979 · HTML: https://mlsanity.com/paper/2106.14979/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Builds on 39 of its 103 references that are in the map.
- First posted on arXiv 28 Jun 2021.

## Builds on (most cited first, 30 of 39)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems](https://mlsanity.com/paper/2005.01643/) (2020)
- [Neural Collaborative Filtering](https://mlsanity.com/paper/1708.05031/) (2017)
- [Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning](https://mlsanity.com/paper/1604.00923/) (2016)
- [Safe and Efficient Off-Policy Reinforcement Learning](https://mlsanity.com/paper/1606.02647/) (2016)
- [Recommendations as Treatments: Debiasing Learning and Evaluation](https://mlsanity.com/paper/1602.05352/) (2016)
- [Taming the Monster: A Fast and Simple Algorithm for Contextual Bandits](https://mlsanity.com/paper/1402.0555/) (2014)
- [Beyond UCB: Optimal and Efficient Contextual Bandits with Regression Oracles](https://mlsanity.com/paper/2002.04926/) (2020)
- [Top-K Off-Policy Correction for a REINFORCE Recommender System](https://mlsanity.com/paper/1812.02353/) (2018)
- [Unbiased Learning-to-Rank with Biased Feedback](https://mlsanity.com/paper/1608.04468/) (2016)
- [Is a Good Representation Sufficient for Sample Efficient Reinforcement Learning?](https://mlsanity.com/paper/1910.03016/) (2019)
- [Learning with Good Feature Representations in Bandits and in RL with a Generative Model](https://mlsanity.com/paper/1911.07676/) (2019)
- [Bypassing the Monster: A Faster and Simpler Optimal Algorithm for Contextual Bandits under Realizability](https://mlsanity.com/paper/2003.12699/) (2020)
- [Multi-Interest Network with Dynamic Routing for Recommendation at Tmall](https://mlsanity.com/paper/1904.08030/) (2019)
- [Off-policy evaluation for slate recommendation](https://mlsanity.com/paper/1605.04812/) (2016)
- [Corralling a Band of Bandit Algorithms](https://mlsanity.com/paper/1612.06246/) (2016)
- [Learning Tree-based Deep Model for Recommender Systems](https://mlsanity.com/paper/1801.02294/) (2018)
- [Practical Contextual Bandits with Regression Oracles](https://mlsanity.com/paper/1803.01088/) (2018)
- [Mostly Exploration-Free Algorithms for Contextual Bandits](https://mlsanity.com/paper/1704.09011/) (2017)
- [A Contextual Bandit Bake-off](https://mlsanity.com/paper/1802.04064/) (2018)
- [Pixie: A System for Recommending 3+ Billion Items to 200+ Million Users in Real-Time](https://mlsanity.com/paper/1711.07601/) (2017)
- [On the Difficulty of Evaluating Baselines: A Study on Recommender Systems](https://mlsanity.com/paper/1905.01395/) (2019)
- [Misspecified Linear Bandits](https://mlsanity.com/paper/1704.06880/) (2017)
- [Estimating Position Bias without Intrusive Interventions](https://mlsanity.com/paper/1812.05161/) (2018)
- [Do Offline Metrics Predict Online Performance in Recommender Systems?](https://mlsanity.com/paper/2011.07931/) (2020)
- [Recommendations and User Agency: The Reachability of Collaboratively-Filtered Information](https://mlsanity.com/paper/1912.10068/) (2019)

## Built on by (most cited first, 14 of 14)

- [A Survey of Large Language Model Empowered Agents for Recommendation and Search: Towards Next-Generation Information Retrieval](https://mlsanity.com/paper/2503.05659/) (2025)
- [Improving Screening Processes via Calibrated Subset Selection](https://mlsanity.com/paper/2202.01147/) (2022)
- [Performative Recommendation: Diversifying Content via Strategic Incentives](https://mlsanity.com/paper/2302.04336/) (2023)
- [Uncertainty Quantification for Fairness in Two-Stage Recommender Systems](https://mlsanity.com/paper/2205.15436/) (2022)
- [Cooperative Retriever and Ranker in Deep Recommenders](https://mlsanity.com/paper/2206.14649/) (2022)
- [GBA: A Tuning-free Approach to Switch between Synchronous and Asynchronous Training for Recommendation Model](https://mlsanity.com/paper/2205.11048/) (2022)
- [When Are Two Lists Better than One?: Benefits and Harms in Joint Decision-making](https://mlsanity.com/paper/2308.11721/) (2023)
- [Efficient Personalized Reranking with Semi-Autoregressive Generation and Online Knowledge Distillation](https://mlsanity.com/paper/2603.07107/) (2026)
- [Two-Stage Neural Contextual Bandits for Personalised News Recommendation](https://mlsanity.com/paper/2206.14648/) (2022)
- [Optimizing Novelty of Top-k Recommendations using Large Language Models and Reinforcement Learning](https://mlsanity.com/paper/2406.14169/) (2024)
- [Position Paper: Why the Shooting in the Dark Method Dominates Recommender Systems Practice; A Call to Abandon Anti-Utopian Thinking](https://mlsanity.com/paper/2402.02152/) (2024)
- [Imagery as Inquiry: Exploring A Multimodal Dataset for Conversational Recommendation](https://mlsanity.com/paper/2405.14142/) (2024)
- [Online Two-Stage Submodular Maximization](https://mlsanity.com/paper/2510.19480/) (2025)
- [Credit-assigned Policy Gradient for Early Stage Retrieval in Two-stage Ranking](https://mlsanity.com/paper/2605.26385/) (2026)

## Abstract (arXiv)

Thanks to their scalability, two-stage recommenders are used by many of today's largest online platforms, including YouTube, LinkedIn, and Pinterest. These systems produce recommendations in two steps: (i) multiple nominators, tuned for low prediction latency, preselect a small subset of candidates from the whole item pool; (ii) a slower but more accurate ranker further narrows down the nominated items, and serves to the user. Despite their popularity, the literature on two-stage recommenders is relatively scarce, and the algorithms are often treated as mere sums of their parts. Such treatment presupposes that the two-stage performance is explained by the behavior of the individual components in isolation. This is not the case: using synthetic and real-world data, we demonstrate that interactions between the ranker and the nominators substantially affect the overall performance. Motivated by these findings, we derive a generalization lower bound which shows that independent nominator training can lead to performance on par with uniformly random recommendations. We find that careful design of item pools, each assigned to a different nominator, alleviates these issues. As manual search for a good pool allocation is difficult, we propose to learn one instead using a Mixture-of-Experts based approach. This significantly improves both precision and recall at K.