# The Bandit's Blind Spot: The Critical Role of User State Representation in Recommender Systems

arXiv: https://arxiv.org/abs/2604.26651 · HTML: https://mlsanity.com/paper/2604.26651/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 47 references that are in the map.
- First posted on arXiv 29 Apr 2026.

## Builds on (most cited first, 19 of 19)

- [Neural Collaborative Filtering](https://mlsanity.com/paper/1708.05031/) (2017)
- [RecVAE: a New Variational Autoencoder for Top-N Recommendations with Implicit Feedback](https://mlsanity.com/paper/1912.11160/) (2019)
- [Frequency Enhanced Hybrid Attention Network for Sequential Recommendation](https://mlsanity.com/paper/2304.09184/) (2023)
- [An Attentive Inductive Bias for Sequential Recommendation beyond the Self-Attention](https://mlsanity.com/paper/2312.10325/) (2023)
- [Deep Bayesian Bandits: Exploring in Online Personalized Recommendations](https://mlsanity.com/paper/2008.00727/) (2020)
- [Graph Neural Bandits](https://mlsanity.com/paper/2308.10808/) (2023)
- [Latent Contextual Bandits and their Application to Personalized Recommendations for New Users](https://mlsanity.com/paper/1604.06743/) (2016)
- [Scalable Neural Contextual Bandit for Recommender Systems](https://mlsanity.com/paper/2306.14834/) (2023)
- [Spoiled for Choice? Personalized Recommendation for Healthcare Decisions: A Multi-Armed Bandit Approach](https://mlsanity.com/paper/2009.06108/) (2020)
- [Diffusion-based Contrastive Learning for Sequential Recommendation](https://mlsanity.com/paper/2405.09369/) (2024)
- [Dynamic Embedding Size Search with Minimum Regret for Streaming Recommender System](https://mlsanity.com/paper/2308.07760/) (2023)
- [Alternating Linear Bandits for Online Matrix-Factorization Recommendation](https://mlsanity.com/paper/1810.09401/) (2018)
- [Bandits Warm-up Cold Recommender Systems](https://mlsanity.com/paper/1407.2806/) (2014)
- [Relational Boosted Bandits](https://mlsanity.com/paper/2012.09220/) (2020)
- [BanditMF: Multi-Armed Bandit Based Matrix Factorization Recommender System](https://mlsanity.com/paper/2106.10898/) (2021)
- [Preferences Evolve And So Should Your Bandits: Bandits with Evolving States for Online Platforms](https://mlsanity.com/paper/2307.11655/) (2023)
- [To Copy, or not to Copy; That is a Critical Issue of the Output Softmax Layer in Neural Sequential Recommenders](https://mlsanity.com/paper/2310.14079/) (2023)
- [Accelerated learning from recommender systems using multi-armed bandit](https://mlsanity.com/paper/1908.06158/) (2019)
- [Uncertainty of Joint Neural Contextual Bandit](https://mlsanity.com/paper/2406.02515/) (2024)

## Abstract (arXiv)

With the increasing availability of online information, recommender systems have become an important tool for many web-based systems. Due to the continuous aspect of recommendation environments, these systems increasingly rely on contextual multi-armed bandits (CMAB) to deliver personalized and real-time suggestions. A critical yet underexplored component in these systems is the representation of user state, which typically encapsulates the user's interaction history and is deeply correlated with the model's decisions and learning. In this paper, we investigate the impact of different embedding-based state representations derived from matrix factorization models on the performance of traditional CMAB algorithms. Our large-scale experiments reveal that variations in state representation can lead to improvements greater than those achieved by changing the bandit algorithm itself. Furthermore, no single embedding or aggregation strategy consistently dominates across datasets, underscoring the need for domain-specific evaluation. These results expose a substantial gap in the literature and emphasize that advancing bandit-based recommender systems requires a holistic approach that prioritizes embedding quality and state construction alongside algorithmic innovation. The source code for our experiments is publicly available on https://github.com/UFSCar-LaSID/bandits_blind_spot.